(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Regions, rescources, and economic growth"

UNIVERSITY 
OF FLORIDA 
LIBRARIES 




Digitized by the Internet Archive 
in 2013 



http://archive.org/details/regionsrescourceOOperl 



Regions, Resources, and Economic Growth 



Regions, 
Resources, and 
Economic 
Growth 




Harvey S. Perloff 
Edgar S. Dunn, Jr. 
Eric E. Lampard 
Richard F. Muth 



published FOR Resources for the Future, Inc. 
b y The Johns Hopkins Press, Baltimore 



© 1960 by The Johns Hopkins Press, Baltimore 18, Md. 

Distributed in Great Britain by Oxford University Press, London 

Printed in the United States of America by The Garamond Press, Baltimore 

Library of Congress Catalog Card Number 60-12311 



RESOURCES FOR THE FUTURE, INC., Washington, D. C. 

Board of Directors, Horace M. Albright, Chairman, E. J. Condon, Joseph L. Fisher, Reuben 
G. Gustavson, Otto H. Liebers, Leslie A. Miller, William S. Paley, Laurance S. 
Rockefeller, Stanley H. Ruttenberg, John W. Vanderwilt, P. F. Watzek 

President and Executive Director, Joseph L. Fisher 

Associate Director, Irving K. Fox 

Secretary-Treasurer, John E. Herbert 



Resources for the Future is a nonprofit corporation for research and education in the 
development, conservation, and use of natural resources. It was established in 1952 with 
the co-operation of The Ford Foundation and its activities since then have been financed 
by grants from that Foundation. Part of the work of Resources for the Future is carried 
out by its resident staff, part supported by grants to universities and other nonprofit 
organizations. Unless otherwise stated, interpretations and conclusions in RFF publications 
are those of the authors; the organization takes responsibility for the selection of significant 
subjects for study, the competence of the researchers and their freedom of inquiry. 

This book is one of RFF's regional studies, which are directed by Harvey S. Perloff. 
The manuscript was edited by Martha Gibbon. 



Staff Editors, Henry Jarrett and Vera W. Dodds 



Preface 



Economic growth in the United States has varied widely among the different 
geographic sections. Some regions have expanded rapidly, drawing in per- 
sons, materials, and capital in great quantities and enjoying high levels of 
living, at the same time that others are in various stages of relative decline. 

The move of the Braves baseball team from Boston to Milwaukee and of 
the Dodgers and Giants across the continent to the West Coast might be said 
to symbolize two of the significant economic shifts of recent decades — the 
particularly rapid industrial growth of the western end of the Manufacturing 
Belt and the virtual flood of movement to the Far West, and especially Cali- 
fornia, from almost every other part of the country. The regional changes are 
not always so dramatic, but they are continuous and they play a large role in 
where and how our families live and in how they prosper. 

This differential growth strongly influences the development and use of 
natural resources which, in turn, have an important effect on regional growth 
patterns. To understand the complicated interrelationships involved, it is neces- 
sary to see the natural resource activities — the farming, forestry, fishing, and 
mining, and the processing of their products — within the broad framework of 
over-all national and differential regional economic growth. The study presented 
here attempts to do this. 

It was designed with two objectives in view: 

One was to furnish information and insights that might be of value to the 
public and private groups concerned with various aspects of economic growth 
and decline as well as with natural resources. This is a field in which action 
programs seem to have grown at a much faster rate than the underlying knowl- 
edge essential to their effectiveness. Over 10,000 public and private agencies 
in the United States are concerned in one way or another with "area develop- 
ment." It has become increasingly evident — particularly to thoughtful persons 
with "area development" responsibilities — that efforts in the various sections 
of the country to attract industry and to raise levels of living can hope to be 
successful only if such policies are based firmly on a knowledge of industry 
requirements and impacts. It was our hope that the study could contribute 
to the knowledge needed for a sound policy approach to regional economic and 
resources planning and development. 

The second objective was to provide a conceptual and methodological frame- 

/ v 



vi / Preface 

work as a guide for our own future research in resources and regions and for 
possible use by scholars working in regional and developmental economics, 
business economics, geography, planning, and related fields, both in the United 
States and in other countries. While the study is concerned exclusively with 
developments in the U.S. economy, the conceptual and methodological prob- 
lems involved in understanding the underlying forces at work have general 
relevance. In evolving our research approach, we have borrowed heavily from 
the work of scholars of many countries and we would hope that this study in 
turn can contribute something of value to those concerned with problems of 
subnational economic growth overseas. 

Both objectives dictated a broad view of the subject, with a focus on 
economic growth as it takes place both over time and in space. Several assump- 
tions growing out of our own previous research and that of others underlie our 
over-all approach: 

(1) That regional economic growth in the volume of economic activities 
is subject to a different constellation of socio-economic forces than are the 
welfare aspects of growth. Both of these aspects received detailed attention in 
the study; all of Part V is devoted to welfare considerations, while Parts III 
and IV are chiefly concerned with a description and analysis of regional 
changes in the volume of economic activities. (2) That economic growth is an 
evolutionary process, with current development and future possibilities greatly 
influenced by the decisions and developments of the past — such as the existing 
size of markets as they have evolved in different regions of the country, the 
location of input sources, the economics of capital depreciation, and related 
considerations. Thus, in our research we sought an understanding of the 
regional settlement and growth patterns of the past as a necessary foundation 
for an understanding of present differential levels of living and rates of 
economic expansion. 

(3) That the critical elements in the changing patterns of regional economic 
development are the locational and production decisions of business firms 
flowing from the input and market requirements of the major industries. This 
called for an overview of the key requirements of each of the major industry 
groups — a large task but, in our view, an essential one. 

As the research unfolded, we found in certain of the initial findings a gold 
mine of knowledge and insight. This was true, for example, of our discovery, 
based on the application of a useful statistical tool, that the regions were 
growing in two distinctive ways: that certain regions were growing largely as a 
result of "within-industry" changes (of the type symbolized by the shift of 
cotton cultivation from the Southeast to the Southwest), while others were 
gaining from their particular industrial composition, that is, because they had 
an unusually high proportion of "rapid-growth" industries. So instructive was 
this distinction that it could be used as a generalized descriptive and analytical 
framework. 

The book is divided into five parts. Part I presents the highlights in the 
story of regional economic growth in the United States since 1870, as well as 



Preface / vii 

a more detailed account of growth during the recent period 1939-54, as back- 
ground for the analysis of the factors behind differential regional growth 
presented in later sections of the book. Part II discusses theories of growth 
and the conceptual and methodological framework employed in the study. Part 
III discusses the evolution of the subnational economy from the early part of the 
19th Century, with a more detailed treatment of changes in the spatial distri- 
bution of population, income, natural resource activities, and other economic 
activities from 1870 to 1950. Part IV analyzes the forces behind the recent 
locational shifts in American industry and the impact of these shifts on differ- 
ential regional economic growth. Part V is devoted to an analysis of the major 
causes of regional differentials in the level of per capita income and in the 
rates of growth in per capita income. 

Summary statements are provided at several points. There is a summary 
of the key facts in regional economic development since 1870 in the last section 
of chapter 3. The emphasis here is on the growth indexes. A roundup of the 
main facts in the development of productive activities in the multistate regions 
of the country, and particularly the natural resources activities, is provided 
in chapter 18 at the end of Part III. A summary of the recent industrial shifts, 
of changes in the regional distribution of economic activities, and of the 
resulting regional patterns of growth is provided in chapters 19 and 26, the 
opening and closing chapters of Part IV. The central features of the regional 
variations in levels of living and of growth of per capita income are summarized 
at the end of each of the chapters in Part V and in more general terms in 
chapter 8 of Part II and chapter 34 of Part V. 

The summary for the entire volume is provided in chapter 9, at the end of 
the conceptual discussion and preceding the detailed description and analysis 
of regional economic growth which are set out in Parts III, IV and V. 

While each of the authors collaborating in this study has had major respon- 
sibility for certain sections of it, each has contributed ideas and materials to 
other parts of the book as well. Thus, the over-all design of the study and its 
direction were my main responsibility, as was the presentation of the intro- 
ductory and conceptual material in Parts I and II, but each of the members 
of the group made important contributions to both of these tasks. 

Eric E. Lampard brought together the historical data for the period 1870- 
1950, which involved the preparation of a number of new or improved historical 
series, and prepared Part III, weaving together many of the strands in the rich 
fabric of regional economic development in the United States. 

Edgar S. Dunn, Jr., prepared Part IV, which analyzes the changes in the 
locational structure of American economic activity in the crucial years during 
and since World War II. This analysis is based on the remarkable feat of 
examining the key locational factors in each of the major industrial groups in 
the U. S. economy. 

Richard F. Muth prepared most of Part V, in which the major causes of 
regional differentials in the level of per capita income and in the rates of 
growth in per capita income are analyzed, and provided valuable assistance 



viii / Preface 

to me in the preparation of chapter 29 of Part V, "Income and the Industrial 
Structure," chapter 33, "Migration and Income," and chapter 34, "Regions 
and Income: a summary." 

Dunn and Lampard each took a year's leave from the University of Florida 
and Smith College respectively to work on the study full time at Resources for 
the Future; and both spent a further year completing their phases of the work 
after returning to their own institutions. During this period Muth was 
associated full-time with the Program of Regional Studies at Resources for the 
Future. 

Throughout the book we have made use of materials focusing on the State 
Economic Areas (groups of counties with more or less common economic and 
social characteristics, as denned by the Bureau of the Census). These materials 
— relating to structure of economic activities, natural and human resources, 
migration, and levels of living — were prepared by Otis Dudley Duncan, who 
directed the work, Ray P. Cuzzort, and Beverly Duncan, under a grant to the 
Population Research and Training Center at the University of Chicago. 

This volume is one of a series of Resources for the Future studies aimed at 
deepening our understanding of the role of natural resources in economic 
growth. These have been undertaken to help provide a foundation for policies 
and practices on the development, use, and conservation of natural resources 
which will contribute most to the general welfare. 

A companion study, under Otis Dudley Duncan's direction, has just been 
published under the title, Metropolis and Region. This study deals specifically 
with the economic structure and functions of metropolitan communities and 
their relation to larger regions and to the nation. The two studies represent a 
loose division of labor between our own concentration on the state and the 
multistate region as a way of examining the subnational economy and the 
focus on the substate areas of the two efforts under Duncan's direction. As 
a result each group could study the subnational economy with fairly consistent 
types of data and thus hope to highlight somewhat different facets of the com- 
plex problems of regional economic structure and growth. 

Attention in this book and in Metropolis and Region is directed to the 
continental United States. A separate study of economic development and 
resource problems in Alaska is being carried out by George W. Rogers under 
a grant from Resources for the Future to the Arctic Institute of North America. 

Our study was undertaken during a period in which there has been a flower- 
ing of regional studies. This has provided a stimulus to our own research; 
it also made it unnecessary for us to cover all the major aspects of the subject 
in comparable detail. Those who can make the best use of material on regional 
economic growth are likely to examine at least the major studies as they become 
available. 

Thus, we were fortunate in having at hand a prepublication copy of the 
University of Pennsylvania study (Kuznets, Thomas, Lee, Miller, Brainerd, 
Easterlin), Population Redistribution and Economic Growth, United States, 
1870-1950, as well as a typescript copy of Frank A. Hanna's book, State Income 
Differentials, 1919-1954. We made extensive use of these materials in our re- 



Preface / ix 

search. At the same time, knowing that these volumes would be available before 
our own, we could touch lightly on the topics covered by these valuable studies 
without any sense of neglect. 

The same was true of a number of other subjects important in the examina- 
tion of regional economic growth. For example, during the period of our 
research we were in close touch with the group, under the direction of Raymond 
Vernon, carrying out the New York Metropolitan Region Study. Because we 
were aware of the very comprehensive treatment being given to problems of 
transportation and to the tertiary (service) activities in this study, which will 
be available at about the same time as our own, we felt that we could devote 
major attention to other aspects of greater direct importance to our central 
theme. 

We gained greatly also from a number of preliminary unpublished studies, as 
well as published materials which were generously provided to us by Walter 
Isard and his associates at the University of Pennsylvania, William Nicholls 
and his associates at Vanderbilt University, and the regional-economics research 
group at Brown University. 

We are very much in debt to the following persons who reviewed the entire 
manuscript and made many valuable suggestion: George Borts, 0. Dudley 
Duncan, Richard A. Easterlin, Jerome L. Stein, Richard S. Thoman, Werner Z. 
Hirsch, and colleagues at Resources for the Future, Marion Clawson, Joseph 
Fisher, Henry Jarrett, and T. Lowdon Wingo, Jr. Wingo enriched his comments 
with new materials which we could use to great advantage in the book. Thoman 
also prepared a number of the charts used in the volume and provided materials 
drawn from the field of economic geography. 

We are also indebted to the following scholars who reviewed one or several 
chapters in the field in which they have special competence or interest: Kenneth 
L. Bachman, Ronald L. Mighell, and R. Burnell Held on the agricultural sec- 
tions; Charles H. Stoddard on forestry; Orris C. Herfindahl, Joseph Lerner, and 
Bruce C. Netschert on energy and mineral resources; Douglass C. North and 
William N. Parker on the historical materials of Part III; Harold J. Barnett, 
Paul H. Cootner, W. Lee Hansen, Douglass C. North, and Charles M. Tiebout 
on Parts I, II, and V. These reviews have contributed a great deal to improving 
the book. 

Extremely able statistical assistance was provided by Mrs. Erna J. Peters, 
while the tremendous task of typing the many drafts of the book was done by, 
and under the supervision of, Mrs. Nadja Schocken. Miss Martha Gibbon did a 
thorough job of editing the volume. We were fortunate in having such com- 
petent assistance. 

Harvey S. Perloff 



Contents 



Part 1. An Historical Review of Regional Economic Growth 

1. Introduction: A note on regions, time, and growth measures . . 3 

2. Economic growth since 1870 (with a focus on the multi-state region) 9 

3. Economic growth, 1939-1954 (with a focus on the states) ... 28 

Part 11. A Framework for Analysis 

4. Theories of regional economic growth 55 

5. Factors influencing the volume and composition of regional economic 

activity 63 

6. Location factors 75 

7. "Input-output access" and multipliers 87 

8. Factors influencing per capita incomes 97 

9. Summary 104 

Part 111. Regional Economic Development, 1870-1950 

10. Regional differentiation in the American economy before 1870 . 109 

11. Changing regional structure of the U.S. economy, 1870-1910: 

(1) growth and shift of population 122 

12. Changing regional structure of the U.S. economy, 1870-1910: 

. (2) shifts in productive activities 130 

13. Changing regional structure of the U.S. economy, 1870-1910: 

(3) regional industrial structures and income 170 

14. Nationalizing the American economy, 1870-1910: the play of tech- 

nology on resources 191 

15. Regional adjustments within the national economy, 1910-1950: 

(1) population shifts 222 

16. Regional adjustments within the national economy, 1910-1950: 

(2) shifts in productive activities 233 

17. Regional adjustments within the national economy, 1910-1950: 

(3) regional industrial structures and income 263 

18. Resource activities and regional economic growth since 1870: a 

retrospect 284 

/ xi 



xii / Contents 

Part IV. The Regional Distribution of Economic Activities 
in the United States, 1939-1954 

19. An approach to the study of changes in the regional distribution of 

economic activities 295 

20. Regional distribution of mining activity 307 

21. Regional distribution of agricultural activity: (1) farming in 

general 335 

22. Regional distribution of agricultural activity: (2) the major com- 

ponents 355 

23. Regional influence of manufacturing activity: (1) manufacturing 

in general 380 

24. Regional influence of manufacturing activity: (2) the major com- 

ponents 408 

25. Regional influence of manufacturing activity: (3) the major com- 

ponents; service activities 442 

26. The resulting regional patterns of growth 467 

Part V. Variations in Levels and Rates of Growth 
of Per Capita Income 

27. Inter-state differences in per capita income and in rates of growth in 

per capita income 489 

28. Rates of income change, 1920-1955 502 

29. Income and industrial structure s . . . 520 

30. Factors related to money income differentials 536 

31. Differences in labor incomes in the same industry: (1) agriculture 552 

32. Differences in labor incomes in the same industry; (2) manufactur- 

ing 568 

33. Migration and income 589 

34. Regions and income: a summary 600 

Statistical appendix 609 

Index 685 



TABLES 



Part I. An Historical Review of Regional Economic Growth 

1. Regional distribution of U. S. population, 1870-1957, and population 

of regions, 1870 and 1957 12 

2. Percentage increases in population by census decades, by region, 

1870-1950 (and 1950-1957) 13 

3. Density of population, by region, 1870-1957 (population per square 

mile of land area) 14 

4. Urban population as percentage of total population, by region, 1870, 

1910, and 1950 19 

5. Decennial rates of increase of urban population by region, 1870-1950 20 

6. Rank order of regions by rate of increase in urban population and 

regional proportions of U.S. urban population, 1870-1950 . . 21 

7. City size, 1870, 1910, and 1950 22 

8. Regional distribution of total personal income, 1880-1957 ... 24 

9. Regional income per capita, 1880-1957, and as percentage of na- 

tional average 27 

10. Principal population shifts among the states, 1910-1957 ... 35 

11. Principal shifts in total personal income among the states, 1920- 

1957 37 

12. Means and standard deviations of farm-operator family level-of- 

living indexes for 348 non-metropolitan state economic areas, 
1930, 1940, 1945, 1950, and 1954 45 



Part II. A Framework for Analysis 

13. An illustration of the differential shift for selected states, 1939-1954 73 

14. Illustration of total employment shifts derived from a combination 

of the differential effect and the proportionality effect for selected 
states, 1939-1954 74 

15. Summary of rank correlations: economic growth and changes in 

employment structure, among the states, 1880-1920 and 1920- 
1950 98 



Part III. Regional Economic Development, 1870-1950 

16. Population, labor force, and income originating in commodity pro- 

duction in the United States, by region, 1840 114 

17. Percentage distribution of U. S. population in the "West," 1840 

and 1860 116 

18. Larger urban centers of the "West" in 1860, ranked in order of 

population engaged in manufactures 118 

19. Leading branches of manufacture in the United States, 1860 . . 119 

/ xiii 



xiv / Contents 

20. Employment in manufactures in the United States, by region, 1850 

and 1860 120 

21. Rank order of regions by personal income per capita, urbanization 

of population, and labor force industry components, 1880 . . 120 

22. Percentage increases in population by census decades, by region, 

1870-1910 123 

23. Density of population, by region, 1870, 1890, and 1910 .... 124 

24. Regional distribution of U. S. population, 1870-1910 .... 124 

25. Principal net shifts in population among states, 1870-1910 . . . 126 

26. Urban population as percentage of total population, by region, 1870 

and 1910 126 

27. Decennial rates of increase of urban population, by region, 1870- 

1910 127 

28. Rank order of regions by rate of increase in urban population, 

1870-1880 and 1900-1910, and regional proportions of U. S. 
urban population, 1870 and 1910 127 

29. Principal net shifts in urban population among states, 1870-1910 . 128 

30. U. S. labor force, distribution by major industry group, 1870, 1890, 

and 1910 131 

31. Regional distribution of U. S. labor force, 1870, 1890, and 1910 . 131 

32. Percentage change in regional shares of U. S. population and labor 

force, 1870-1910 132 

33. Regional distribution of all U. S. resource industries labor force, 

1870, 1890, and 1910 133 

34. Regional distribution of each resource industry's labor force 

1870-1910 134 

35. Percentage distribution of each resource industry's employment, 

by great region, 1870 and 1910 136 

36. Regional distribution of value of primary U. S. resources extracted, 

1870, 1890, and 1910 . 137 

37. Distribution of labor force in resource industries, and value of 

extracted resources, by great region, 1870, 1890, and 1910 . . 138 

38. Regional distribution of value of resources extracted, by major 

resource industry, 1870, 1890, and 1910 138 

39. Percentage distribution of value of all resources extracted, by major 

resource industry, 1870 and 1910 140 

40. Percentage distribution of value of all resources extracted, by major 

resource industry and great region, 1870, 1890, and 1910 . . 141 

41. Principal net shifts in agricultural labor force among states, by 

decade, 1870-1910 143 

42. Principal net shifts in mining labor force among states, by decade, 

1870-1910 146 

43. Principal net shifts in forest labor force among states, by decade, 

1870-1910 148 

44. Regional distribution of the manufacturing labor force, 1870, 1890, 

and 1910 152 



Tables / xv 

45. Distribution of manufacturing labor force, by great region, 

1870-1950 152 

46. Regional distribution of U. S. value added by manufacture, 1870, 

1890, and 1910 153 

47. Distribution of manufacturing labor force and value added by manu- 

facture, by great region, 1870, 1890, and 1910 153 

48. Principal net shifts in manufacturing labor force among states, 

1870-1910 156 

49. Regional distribution of employment in ten selected manufacturing 

industries, 1870, 1910, and 1954 158 

50. Regional distribution of services labor force, 1870, 1890, and 1910 162 

51. Percentage change in regional shares of population, total labor force, 

and services labor force, 1870-1910 163 

52. Principal net shifts in services labor force among states, 1870-1910 163 

53. Percentage growth in total labor force, services labor force, and 

labor force in four selected service industries, 1870-1910 . . 164 

54. Regional distribution of labor force in four selected service 

industries, 1870 and 1910 165 

55. Labor force in four selected service industries as a percentage of 

total labor force, by region, 1870, 1910, and 1950 .... 167 

56. Percentage change in regional distribution of labor force in four 

selected service industries, 1870-1910 168 

57. U. S. labor force, distribution by industry, 1870-1910 .... 172 

58. Distribution by industry of New England region labor force, 

1870-1910 172 

59. Distribution by industry of Middle Atlantic region labor force, 

1870-1910 173 

60. Distribution by industry of Great Lakes region labor force, 

1870-1910 174 

61. Distribution by industry of Southeast region labor force, 1870-1910 176 

62. Distribution by industry of Plains region labor force, 1870-1910 . 178 

63. Distribution by industry of Southwest region labor force, 1870-1910 179 

64. Distribution by industry of Mountain region labor force, 1870-1910 181 

65. Distribution by industry of Far West region labor force, 1870-1910 182 

66. Combined manufacturing-services components of labor force of Far 

West, New England, Middle Atlantic, and Great Lakes regions, 
1870, 1890, and 1910 183 

67. Regional distribution of total personal income in the United States, 

1880, 1900, and 1920 185 

68. Principal net shifts in total personal income among states, 1880- 

1920 185 

69. Rank order of regions by personal income per capita, 1880, 1900, 

and 1920 187 

70. Rank order of regions by personal income per capita, urbanization 

of population, and labor force industry components, 1920 . , 188 



xvi / Contents 

71. Resource and non-resource components of the labor force in the 

Mountain and Far West regions, 1900 and 1920 190 

72. Railroad mileage in operation during calendar years and percentage 

increases by decades, by region, 1370-1910 196 

73. Regional distribution of selected farm products, 1370-1950 . . 193 

74. Leading states in the production of whole milk sold from farms, 

wheat threshed, and all cattle, 1370-1950, as a percentage of U. S. 

total 203 

75. Domestic mine production of recoverable lead, by region, 1870-1950 205 

76. Domestic copper mine production, by region, 1870-1950 . . . 208 

77. Pig iron production, by region, 1870-1950 210 

78. (a) Coke production, by region, 1880-1950; (b) Coal production, 

by region, 1870-1950 212 

79. Percentage increases in population by census decades, by region, 

1910-1950 223 

80. Density of population, by region, 1910, 1930, and 1950 .... 224 

81. Regional distribution of U. S. population, 1910-1950 .... 225 

82. Principal net shifts in population among states, 1910-1950 . . 225 

83. Urban population as percentage of total population, by region, 1910 

and 1950 227 

84. Decennial rates of increase of urban population, by region, 

1910-1950 229 

85. Rank order of regions by rate of increase in urban population, 

1900-1910 and 1940-1950, and regional proportions of U. S. 
urban population, 1910 and 1950 229 

86. Principal net shifts in urban population among states, 1910-1950 . 230 

87. Regional distribution of cities in the United States with 50,000 or 

more inhabitants, by size group, 1950 231 

83. U. S. labor force, distribution by major industry group, 1910, 1930, 

and 1950 234 

89. Rank order of states with greatest relative concentration of foreign 

born white and of foreign white stock, 1920 and 1950 . . . 235 

90. Regional distribution of U. S. labor force, 1910, 1930, and 1950 . 236 

91. Percentage change in regional shares of U. S. population and labor 

force, 1910-1950 237 

92. Regional distribution of all U. S. resource industries labor force, 

1910, 1930, and 1950 238 

93. Regional distribution of each resource industry's labor force, 

1910-1950 238 

94. Principal net shifts in agricultural labor force among states, 

1910-1950 241 

95. Principal net shifts in mining labor force among states, 1910-1950 242 

96. Principal net shifts in forest labor force among states, 1910-1950 . 245 

97. Principal net shifts in fishing labor force among states, 1910-1950 246 

98. Regional distribution of value of primary resources extracted, 1910, 

1930, and 1950 . 247 



Tables / xvii 

99. Percentage distribution of value of all resources extracted, by major 

resource industry and great region, 1910, 1930, and 1950 . . 248 

100. Regional distribution of value of resources extracted, by major 

resource industry, 1910, 1930, and 1950 249 

101. Distribution of labor force in resource industries and value of 

extracted resources, by great region, 1910, 1930, and 1950 . . 251 

102. Regional distribution of manufacturing labor force, 1910, 1930, 

and 1950 252 

103. Regional distribution of value added by manufacture, 1910, 1930, 

and 1950 252 

104. Distribution of manufacturing labor force and value added by 

manufacture, by great region, 1910, 1930, and 1950 .... 253 

105. Principal net shifts in manufacturing labor force among states, 

1910-1950 255 

106. Principal net shifts in manufacturing labor force among states, 

by decade, 1910-1950 256 

107. Regional distribution of services labor force, 1910, 1930, and 1950 257 

108. Percentage change in regional shares of population, total labor force, 

and services labor force, 1910-1950 258 

109. Principal net shifts in services labor force among states, 1910-1950 259 

110. Percentage growth in total labor force, services labor force, and 

labor force in four selected service industries, 1910-1950 . . 259 

111. Regional distribution of labor force in four selected service 

industries, 1910 and 1950 260 

112. Labor force in four selected service industries as a percentage of the 

total labor force, by regions, 1910 and 1950 261 

113. Percentage change in trade and finance shares of regional and total 

labor forces, 1910-1950 262 

114. Labor force, distribution by industry, 1910-1950 264 

115. Distribution by industry of New England region labor force, 

1910-1950 264 

116. Distribution by industry of Middle Atlantic region labor force, 

1910-1950 265 

117. Distribution by industry of Great Lakes region labor force, 

1910-1950 266 

118. Distribution by industry of Southeast region labor force, 1910-1950 267 

119. Distribution by industry of Plains region labor force, 1910-1950 . 269 

120. Distribution by industry of Southwest region labor force, 1910-1950 270 

121. Distribution by industry of Mountain region labor force, 1910-1950 271 

122. Distribution by industry of Far West region labor force, 1910-1950 272 

123. Regional distribution of total personal income in the United States, 

1920-1950 274 

124. Principal net shifts in total personal income among states, 1920-1950 275 

125. Rank order of regions by personal income per capita, 1920-1950 . 276 

126. Regional peaks of population and total personal income shares, and 



xviii / Contents 

trend of regional relative to national per capita income in the 
United States since 1870 278 

127. Rank order of regions, by personal income per capita, urbanization 

of population, and labor force industry components, 1950 . . 282 

Part IV. The Regional Distribution of Economic Activities 
in the United States, 1939-1954 

128. Relative importance of employment shifts, within the major employ- 

ment sectors, 1939-1954 300 

129. Total, differential, and proportionality net shifts in employment, 

by state, 1939-1954 301 

130. Percentage change in employment in the United States, by industry, 

1939-1954 303 

131. Relative employment in rapid-growth and slow-growth industries in 

the United States, by state, 1939 304 

132. Relative importance of net employment shifts in mining, 1939-1954 308 

133. Mining employment as percentage of total employment, 1939, for 

states with net downward proportionality shifts in total employ- 
ment, 1939-1954, because of specialization in mining . . . 309 

134. States with relatively largest net differential shifts in mining employ- 

ment, 1939-1954 312 

135. Percentage change in mining employment in the United States by 

2-digit industry group, 1939-1954 314 

136. Rank correlations of each state's share of employment in selected 

industries with each state's share of population, 1954 . . . 315 

137. Relative use of coal and coke and of oil and natural gas in manu- 

facturing in Pennsylvania, West Virginia, and Texas, by industry, 
1947 324 

138. Rank correlations of each state's share of employment in selected 

mining activities with each state's share of population, 1954 . . 329 

139. Agricultural employment as percentage of total employment, 1939 

and 1954, for states with net downward proportionality shifts in 
total employment, 1939-1954, because of specialization in 
agriculture . 337 

140. Differential shifts in agricultural employment as percentage of total 

employment for states with largest net differential shifts in 
agricultural employment, 1939-1954 342 

141. Relative importance of net shifts in value of agricultural products 

sold, 1940-1954 345 

142. Percentage change in value of agricultural products sold for 2-digit 

agricultural sectors, 1940-1954 346 

143. Rank correlations of each state's share of selected agricultural 

activities with each state's share of population, 1939 and 1954 . 348 

144. Percentage distribution of inputs and outputs for major industry 

groups, 1947 351 



Tables / xix 

145. Manufacturing employment as percentage of total employment in 

states experiencing net upward proportionality shifts mainly 
because of specialization in manufacturing, 1939 and 1954 . . 386 

146. Differential shifts in manufacturing employment as percentage of 

total shifts in employment, 1939-1954 383 

147. Percentage change in manufacturing production worker employment 

for 2-digit manufacturing sectors, 1939-1954 391 

148. Rank correlations of selected resource and resource-using manu- 

facturing groups, 1954 394 

149. Per cent of inputs coming from manufacturing and per cent of 

outputs going to manufacturing for selected 2-digit manufacturing 
categories, 1947 396 

150. Rank correlations with population of manufacturing employment in 

selected 2-digit groups, 1954, based upon state data .... 398 

151. 1-digit total, differential, and proportionality net shifts in manu- 

facturing employment and the 2-digit components of the 1-digit 
differential effect for each state, 1939-1954 399 

152. Relative importance of net employment shifts in manufacturing 

production workers, 1939-1954 409 

153. Net shifts in textile manufacturing employment, selected regions, 

1939-1954 418 

154. Percentage net shift in employment in plastic materials and plastic 

products, by region, 1947-1954 436 

155. Distribution by district of steel-making capacity in the United States, 

1930-1954 450 

156. Aluminum plants of the United States, June 1956 453 

157. State total, differential, and proportionality net shifts in employment 

and the industry components of the differential shifts, 1939-1954 468 

158. Employment shifts in California, by industry, 1939-1954 . . . 472 

159. Manufacturing employment shifts in California, by manufacturing 

industry group, 1939-1954 474 

160. Employment shifts in Texas, by industry, 1939-1954 .... 477 

Part V. Variations in Levels and Rates of Growth 
of Per Capita Income 

161. Per capita personal income and its components, by region, selected 

states, 1950 . . . 494 

162. Median income of families and unrelated individuals, by regions, 

selected states, 1949 496 

163. Median income, persons with income, 14 years old and over, by 

region, selected states, 1949 498 

164. Median income, white males with income, 14 years old and over, 

standard metropolitan areas, by size, 1949 499 

165. Median income, Negro males with income, 14 years old and over, 

standard metropolitan areas, by size, 1949 500 



xx / Contents 

166. Per capita personal income, by region, selected states, 1950 dollars 

1929-1955 504 

167. Per capita income from current production, participation plus 

property income, by region, selected states, selected years, 
1920-1955 507 

168. Per capita participation and property income, by region, selected 

states, 1920 and 1950 509 

169. Participation income per worker, all industry, agricultural and non- 

agricultural, by region, selected states, 1920 and 1950 . . . 512 

170. Per cent of labor force employed in agriculture, by region, selected 

states, 1920 and 1950 514 

171. Per cent change in participation income per worker, by components 

of change, by region, selected states, 1920-1950 516 

172. Arithmetic means of percentage share of three major sectors in 

labor force, groups of states by total income per capita, selected 
years, 1920-1950 521 

173. Industrial distribution of employment, by region, selected states, 

1950 . 524 

174. Rank correlations of income and the relative importance of various 

employment categories, by state, 1950 528 

175. Observed and "expected" income per capita (income-payments 

basis), 1950, on basis of industrial structure 530 

176. Average income per full-time employed person in selected industries, 

1949-1951 532 

177. Changes in percentage shares of three major sectors in labor force, 

1920-1950, groups of states by rise in total income per capita 
over the period 533 

178. Rank correlation coefficients of changes in per capita income and 

industrial structure of the states, 1920-1950 533 

179. Median family and individual incomes, non-farm vs. farm, by 

region, 1949 539 

180. B.L.S. city worker's budget, four regions, October 1949 . . . 540 

181. Reported earnings and Hanna's "rate-constant" occupational and 

industrial earnings, all wage and salary workers, by region, 
selected states, 1949 544 

182. Regional farm wage rate indices, average, 1910-1942, and correc- 

tions for equalizing differences 550 

183. Output per worker in agriculture, by region, selected states, 1950 . 557 

184. Summary of the relation of output per worker and the value of land 

and buildings per worker, by states, 1950 559 

185. Comparison of value of land and buildings per worker, by type of 

farm, with agricultural output per worker, by states, 1950 . . 560 

186. Agricultural labor incomes 562 

187. Median income, rural farm persons . 563 

188. Changes in agricultural output per worker, by region, selected 

states, 1949 U. S. dollars, 1930-1954 566 



Tables / xxi 

189. Covariance analysis of changes in agricultural output per worker . 567 

190. Reported and Hanna's "rate-constant" manufacturing production 

worker hourly earnings, by region, selected states, 1947 . . . 569 

191. Employment in high-capital per labor ratio manufacturing, by 

region, selected states, 1950 574 

192. Relation of the manufacturing wage rate (W) to the per cent of 

manufacturing employment in high capital per labor ratio 
industries (C) by and within regions, 1950 575 

193. Average hourly earnings of manufacturing production workers, by 

industry, by region, 1947 577 

194. Comparison of wages per worker, capital per worker, and value 

added minus wages per worker, all manufacturing, by states, 1919 579 

195. Summary of the relation between wages per production worker man- 

hour and value added less wages per production worker man-hour, 
manufacturing industries, by states, 1947 581 

196. Changes in wages of manufacturing production workers, by region, 

selected states, 1947 U. S. dollars, 1919-1954 583 

197. Covariance analysis of changes in manufacturing production worker 

wages 584 

198. Regional real income per capita, 1880-1950, weighted average of 

state incomes per capita, 1929 dollars 591 

199. Net regional migration per 1000 average population, 1870-1950 . 593 

200. Illustrative migration rates, expected and observed 596 



STATISTICAL APPENDIX 

Source notes 611 

Table 

A1-A7 Labor force, 1870-1950 622 

B1-B5 Value of extractive resources, 1870-1950 637 

C Net shift by states in selected labor force components, 1870-1910 . 642 

D Net shift by states in selected labor force components, 1910-1950 . 646 
E Per capita personal income and its components, 1880, 1900, 1920, 

1940, 1950, and 1957 650 

Fl Total employment and its one-digit components, 1939 .... 652 

F2 Total employment and its one-digit components, 1954 .... 654 
G Percentage increases in employment in the United States in 1-, 2-, 

and 3-digit groups, 1939-1954 656 

H Net shift by states in population, total personal income, total em- 
ployment, and one-digit employment categories, 1939-1954 . . 658 
I Net shift by states in 2-digit mining employment, 1939-1954 . . C65 
J Net shift by states in value of agricultural products sold, 1910-1954 667 



xxii / Contents 

K Net shift by states in 2-digit manufacturing production worker em- 
ployment, 1939-1954 670 

L Important input-output relationships for primary resource extractors 

and 1st and 2nd stage resource users, 1947 677 

M Actual and per cent change, participation (wage and proprietor's) 

income, by components, 1920-1950 681 

N Output per worker and value of land and buildings per worker in 

agriculture, 1950 681 



FIGURES 



Part I. An Historical Review of Regional Economic Growth 

1. Regional grouping of states 5 

2. Population at census dates, by region, 1870-1950 (and 1957) . . 10 

3. Decennial rates of population increase, 1870-1950 (and 1950-1957) 11 

4. Regional distribution of total population, 1870-1950 (and 1957) . 15 

5. Distribution of the population, 1900 16 

6. Distribution of the population, 1950 18 

7. Per capita personal income, 1957 25 

8. Relative state economic growth, 1939-1954 30 

9. Relative regional economic growth, 1939-1954 31 

10. Net shift in total population, 1939-1954 34 

11. Absolute increments to real per capita income, 1939-1954 ... 38 

12. Profile of growth dimensions, by state, 1939-1954 39 

13. Median family income, by state economic areas, 1949 .... 42 

14. Regional patterns of family income (1949) and population growth 

(1940-1950) for SMA's of 100,000 or more population ... 44 

15. Farm-operator family level-of-living indexes for nonmetropolitan 

areas, 1930 46 

16. Farm-operator family level-of-living indexes for nonmetropolitan 

areas, 1954 48 

17. The manufacturing belt, 1900 and 1957 49 



Part II. A Framework for Analysis 

18. Percentage of manufacturing employment in the more rapidly 

growing industries, by state, 1950 69 

19. Increase in manufacturing employment in the more rapidly growing 

industries, as a percentage of the increase in all manufacturing 
employment, 1939-1954 70 

20. Returns on capital in three regions 89 

21. A schematic presentation of types of regions that can exhibit different 

potentials with respect to growth 91 



Figures / xxiii 



Part 111. Regional Economic Development, 1870-1950 



22. Net shift in total population, 1870-1910 125 

23. Net shift in urban population, 1870-1910 128 

24. Net shift in agricultural labor force, 1870-1910 142 

25. Net shift in mining labor force, 1870-1910 145 

26. Net shift in forest labor force, 1870-1910 149 

27. Net shift in fishing labor force, 1870-1910 150 

28. Regional distribution of U. S. manufacturing labor force and value 

added by manufacture, 1870, 1890, and 1910 154 

29. Net shift in manufacturing labor force, 1870-1910 155 

30. Change in the regional and national agricultural labor force as per- 

centage of total labor force, 1870-1910 171 

31. Net shift in total personal income, 1880-1920 186 

32. Regional shares of production, 1870, 1910, 1950 200 

33. Net shift in total population, 1910-1950 226 

34. Net shift in urban population, 1910-1950 228 

35. Net shift in agricultural labor force, 1910-1950 241 

36. Net shift in mining labor force, 1910-1950 243 

37. Net shift in forest labor force, 1910-1950 244 

38. Net shift in fishing labor force, 1910-1950 246 

39. Regional distribution of U. S. manufacturing labor force and value 

added by manufacture, 1930 and 1950 253 

40. Net shift in manufacturing labor force, 1910-1950 254 

41. Net shift in total personal income, 1920-1950 275 

42. The changing regional structure of the labor force, 1870-1950 . . 285 

43. National and regional employment in the primary industries, 1870, 

1910, and 1950 288 

44. Regional proportions of value of primary indexes extracted, 1870- 

1950 289 



Part IV. The Regional Distribution of Economic Activities 
in the United States, 1939-1954 

45. Net shift in total employment, 1939-1954 296 

46. Differential net shift in total employment, 1939-1954 .... 299 

47. Proportionality net shift in total employment, 1939-1954 . . . 299 

48. Net shift in mining employment, 1939-1954 311 

49. Differential net shift in all mining employment, 1939-1954 . . . 313 

50. Proportionality net shift in all mining employment, 1939-1954 . . 313 

51. Net shift in coal mining employment, 1939-1954 317 

52. Tons of coal produced per million tons of reserves, measured in 

British Thermal Units, 1949, by state economic areas .... 320 

53. Relative use of coal and of oil or natural gas by steam electric plants, 

1955, by state 323 



xxiv / Contents 

54. Net shift in employment in crude petroleum and natural gas pro- 

duction, 1939-1954 326 

55. Net shift in metal mining employment, 1939—1954 328 

56. Net shift in non-metallic (other than fuels) mining employment, 

1939-1954 332 

57. Major types of farming in the United States 338 

58. Net shift in agricultural employment, 1939-1954 341 

59. Net shift in value of all agricultural products sold, 1940-1954 . . 343 

60. Total proportionality net shift in value of agricultural products sold, 

1940-1954 347 

61. Total differential net shift in value of agricultural products sold, 

1940-1954 347 

62. Net shift in value of field crops sold, 1940-1954 357 

63. Net shift in value of livestock products sold, 1940-1954 .... 362 

64. Net shift in value of dairy products sold, 1940-1954 365 

65. Net shift in value of poultry products sold, 1940-1954 .... 367 

66. Net shift in value of fruits and nuts sold, 1940-1954 .... 369 

67. Net shift in value of vegetables sold, 1940-1954 369 

68. Net shift in logging employment, 1940-1950 371 

69. Net shift in fishing employment, 1940-1950 375 

70. Per cent of employed labor force in manufacturing, 1950, state 

economic areas 381 

71. Type of manufacturing specialization, 1950, state economic areas . 383 

72. Net shift in total manufacturing employment, 1939-1954 ... 384 

73. Total differential net shift in all manufacturing employment, 

1939-1954 389 

74. Total proportionality net shift in all manufacturing employment, 

1939-1954 390 

75. Net shift in food and kindred products employment, 1939-1954 . 411 

76. Net shift in textile mill employment, 1939-1954 417 

77. Net shift in apparel and related products employment, 1939-1954 . 421 

78. Net shift in lumber and products employment, 1939-1954 . . . 425 

79. Net shift in furniture and fixtures employment, 1939-1954 . . . 428 

80. Net shift in paper and allied products employment, 1939-1954 . . 430 

81. Net shift in chemicals and allied products employment, 1939-1954 433 

82. Net shift in leather and leather products employment, 1939-1954 . 444 

83. Net shift in primary metal industries employment, 1939—1954 . . 449 

84. Net shift in metal products industries employment, 1939-1954 . . 455 

85. Net shift in trade, service, and government activities employment, 

1939-1954 464 



Part V. Variations in Levels and Rates of Growth 
of Per Capita Income 

86. Per capita personal income, 1950 492 

87. Absolute increases in real per capita personal income, 1929-1955 . 505 



Figures / xxv 

88. Increase in participation income per worker due to increase in agri- 

cultural participation income per worker, 1920-1950 .... 517 

89. Increase in participation income per worker due to increase in non- 

agricultural participation income per worker, 1920-1950 . . . 517 

90. Increase in participation income per worker due to fall in % labor 

force employed in agriculture, 1920-1950 518 

91. % manufacturing labor force employed in processing industries, 1950 525 

92. % labor force employed in business service industries, 1950 . . 525 

93. Reported earnings minus industry rate-constant earnings, 1949 . . 546 

94. Output per worker in agriculture, 1950 555 

95. Illustration of covariance analysis: output per worker adjusted for 

capital per worker 558 

96. % manufacturing employment in high capital-labor-ratio industries, 

1950 573 



Part 

J Historical Review of 

Regional Economic Growth 



r> 



7 / Introduction: a note on regions, 
time, and growth measures 



National forces clearly dominate economic developments throughout the whole 
of the United States. Yet average levels of living vary significantly from one 
part of the country to another, and the various states and regions differ from 
one another widely in the size and in the rate of growth of their populations 
and in the scope of their economic activities. The main purpose of our study 
is to arrive at a better understanding of this phenomenon. 

Here in Part I we shall bring together the key facts relevant to our inquiry, 
telling in broad outline the story of regional growth within the context of 
national economic growth in the United States since 1870. But first we shall 
briefly define the growth measures adopted, the regional groupings followed, 
and the time element employed throughout the study in our examination and 
presentation of the data that reflect the phenomenon with which we are con- 
cerned. 



Measures of "Economic Growth" 

Economic growth can be — and is — conceived of in many ways. Inherent 
in the very use of the term is some sense of significant changes in the way in 
which a people produce and consume, work and live and play. To record and 
measure a vast mosaic of change in the nation as a whole and in its various 
parts is no simple undertaking, and it is the better part of wisdom to appreciate 
that, even with our most advanced conceptual and statistical tools, we can grasp 
only the crudest notion of the nature and direction of these changes. 

A useful starting point is to make a distinction between the changes that 
would seem to be associated with individual and family welfare and those 
associated with the volume of economic activities. The most commonly 
employed measures — really, crude indicators — of economic welfare, of improve- 
ment or decline in the average economic status of families and individuals, 
are the relative levels of per capita income and the changes in these levels. 
Different measures are needed in evaluating growth or decline in the volume 
of economic activities — the "more and bigger" aspect of economic growth as 

/ 3 



4 / Historical Review of Regional Economic Growth 

against the "better" aspect. Regional growth in volume might appropriately 
be measured by increases in population (i.e., number of persons, viewed as 
consumers and labor force), increases in total employment, and/or increases in 
total income produced or received within a given area. This distinction between 
the "welfare" measures and the "volume" measures of growth is important 
because it is possible for an area to have an increase in one without a cor- 
responding increase in the other. In other words, an area may have an increase 
in population without an increase in average real per capita income; or an area 
may have a decrease in the volume of economic activities and population and 
yet enjoy an increase in average levels of living. This question is discussed in 
some detail at various points in this book. 

Even when attention is focused on volume alone, the different measures of 
regional growth that might be used provide rather different results. Thus, 
somewhat different patterns may emerge if growth is measured by absolute 
increases in population and employment or by relative increases in population 
and employment, by absolute increases in total income or by relative increases 
in total income. To obtain anything resembling a rounded picture of regional 
economic growth, it is necessary to observe all of these various facets of 
expansion. 1 Also, somewhat different growth patterns may emerge when only 
the one or two most recent decades are considered and when a longer historical 
period is examined. For that reason, we shall examine both. 



Concept of the Region 

The term "region" is generally used to describe a group of geographically 
contiguous areas which have certain common or complementary characteristics 
or which are tied by extensive interareal activity or flows. 2 In economic 
analysis the investigator's choice of component areas (the "building blocks") 
depends both upon the extent to which such areas can be combined in terms of 
specified physical, socio-economic, or other criteria and on the form in which 
statistical data are available or can be made available for the purpose involved. 
This latter element imposes some serious constraints on choice. 

For most of the subjects that are central to the problem to which we have 

lr rhe measurement of economic growth is discussed in some detail in Harvey S. Perloff, 
"Problems of Assessing Regional Economic Progress," Regional Income, National Bureau of 
Economic Research, Studies in Income and Wealth, Vol. 21 (Princeton: Princeton Univer- 
sity Press, 1957), pp. 35-62. See also in the same volume discussions of various aspects of 
the measurement problem by Werner Hochwald, "Conceptual Issues of Regional Income 
Estimation," by Frank Hanna, "Analysis of Interstate Income Differentials: Theory and 
Practice," and by Abner Hurwitz and Carlyle P. Stallings, "Interregional Differentials in 
Per Capita Real Income Change." 

2 For a discussion of the various types of regions employed in economic analysis, see Joseph 
L. Fisher, "Concepts in Regional Economic Development," Papers and Proceedings of the 
Regional Science Association, Vol. 1 (1955), pp. W2-W20. Problems of spatial aggregation 
as related to analysis of trade patterns are discussed by Leon Moses, "The Stability of 
Interregional Trading Patterns and Input-Output Analysis," American Economic Review. 
Vol. 45 (December 1955), pp. 803-32. 



A Note on Regions, Time, and Growth Measures / 5 

addressed ourselves — such as income, natural resource activities, and industrial 
structure — data, and particularly historical series, are available only for the 
states. In several phases of our study — for example, in the discussion of 
industrial location — the subnational picture is drawn mainly in terms of the 
state units. At other points, we have focused on a "regional" grouping of states, 
as useful for both descriptive and analytical purposes. Thus, for example, in 
the historical section, Part III, we have adopted a consistent grouping of states 
as the only practical basis for statistical comparison over long periods of time 
and wide ranges of territory. The grouping is the one followed by the U.S. 
Department of Commerce in presenting its state personal income estimates. 
This was found to be appropriate for our purposes because of its rationale 
with regard to relative levels of living as among the various parts of the country 
and with regard to similarity of cultural-historical backgrounds. That is, this 
grouping follows the main tendencies of historical geographical development 
without doing grave injustice to either local sentiments or chronology of 
settlement. 

A regional grouping of states was also found to be useful in the discussion 
in Part V of per capita income (actually the same grouping as in the historical 
section with one exception). In presenting data which describe the inter-state 
differences in income and its components, averages of income measures for 
regions or groups of states permit an economy in presentation and an easier 
comprehension of inter-state differences. In analyzing these inter-state differ- 
ences, the use of regions provides a check on the completeness of the analysis. 
If the variables which account for the variation of the regional means about 
the national average fail to account for the variation of income measures within 
regions, we have omitted some significant variables from the analysis. 




Figure 1. Regional Grouping of States. 



6 / Historical Review of Regional Economic Growth 



Regional Grouping of States Employed* 

The regional grouping of states that we have adopted for study and presen- 
tation purposes is as follows: 

new ENGLAND: Maine, New Hampshire, Vermont, Massachusetts, Rhode 
Island, Connecticut 

middle ATLANTIC: New York, New Jersey, Pennsylvania, Delaware, Mary- 
land, District of Columbia 

great lakes: Ohio, Indiana, Illinois, Michigan, Wisconsin 

southeast: Virginia, West Virginia, Kentucky, Tennessee, North Carolina, 
South Carolina, Georgia, Florida, Alabama, Mississippi, Arkansas, 
Louisiana 4 

plains: Minnesota, Iowa, Missouri, North Dakota, South Dakota, Nebraska, 
Kansas 

southwest : Oklahoma, Texas, Arizona, New Mexico 

mountain : Montana, Idaho, Wyoming, Utah, Colorado 

far west : Washington, Oregon, California, Nevada 

For purposes of certain broad comparisons, we have grouped these eight 
regions into three "great regions," divided roughly along the courses of the 
Mississippi and Ohio rivers as follows : 

the northeast: embracing the New England, Middle Atlantic, and Great 
Lakes regions, and roughly comparable to the Manufacturing Belt. 

the southeast: embracing the same area as the Southeast region above. 

the west: embracing the Plains, Mountain, Southwest, and Far West 
regions. 

Thus the two great regions, the Northeast and the Southeast, may be con- 
veniently termed the "eastern half" of the nation and the West the "western 

3 This grouping is one that was evolved by a committee composed of representatives of 
three Department of Commerce agencies, convened to consider the feasibility of standardiz- 
ing the regional presentation of statistical data within the Department. The committee 
proposed a nine-region classification of geographically contiguous states after studying a 
large number of statistical series reflecting economic factors, particularly income and 
industrial employment, as well as selected noneconomic factors. We have departed from 
the proposed nine-region grouping by combining the committee's Upper and Lower South 
into one region, giving us eight instead of nine. 

For a report on the Commerce committee's work, see Morris B. Ullman and Robert C. 
Klove, "The Geographic Area in Regional Economic Research," in Regional Income, 
National Bureau of Economic Research, Studies in Income and Wealth, Vol. 21 (Princeton: 
Princeton University Press, 1957), pp. 87-109. For the historical use of the terms "section," 
"division," "region," etc., see articles by F. Mood and V. Carstensen in M. Jensen, ed., 
Regionalism in America (Madison: University of Wisconsin Press, 1952), pp. 5-118. See 
also the revealing treatment of the subject in Otis Dudley Duncan, Ray P. Cuzzort and 
Beverly Duncan, Statistical Geography: Problems in Analyzing Areal Data (Glencoe, HI.: 
Free Press, in press). 

4 For reasons noted later in the text, Louisiana will sometimes be classified with the 
Southwest rather than the Southeast. 



A Note on Regions, Time, and Growth Measures / 7 

half." Fortunately there have been few changes in state or territorial boundaries 
since 1870, with the exception of the division of the Dakota Territory and the 
Indian Territory (mostly Oklahoma) ; hence there has been little necessity 
for the retabulation of census data to accord with our chosen regional designa- 
tion. 5 J 

For us, regional groupings are a device for study and presentation. Unlike 
the statistical agencies, which are necessarily concerned with general-purpose 
groupings that can be employed consistently for long periods of time, we are as 
much concerned with the intra-regional differences as with the similarities, 
with changes in areal patterning over time as with non-changing areal arrange- 
ments. Thus, to take an example, we are as much concerned with the aspects 
in which Florida departs significantly from other states in the Southeast, e.g., 
in income and migration, as with the aspects that distinguish the Southeast as 
a whole from the rest of the country. As a matter of fact, we employ the 
grouping described above merely as a point of departure and feel free to change 
the designation wherever this seems appropriate. Thus, in discussing certain 
subjects, we shall refer to the western end of the Manufacturing Belt, or upper 
New England, or the Atlantic coast states when it is useful to do so. 

In the case of Louisiana, we found in studying economic development in the 
most recent period — over the past two or three decades — that this state had more 
characteristics in common with the Southwest than with the Southeast (particu- 
larly in the growth of the petroleum and oil-using industries). For this reason, 
we have included it with the Southwest in discussions of recent developments in 
Parts IV and V. Also this served as a useful self-discipline, as well as a way of 
alerting readers not to read more into a given regional grouping than the actual 
facts justify. The danger of coming to use a given grouping as more or less a 
final designation, rather than an initial point of reference, is so great that the 
introduction of this type of "change in mid-stream" seems not only justified 
but essential to focus attention on an important matter. 

For certain subjects covered by our study, variations within the states as well 
as among states are analytically significant. We have attempted to get at such 
variations, where particularly significant, by focusing attention on State 
Economic Areas, which are groups of counties and are made up of metropolitan 
and non-metropolitan areas. (The metropolitan SEA is essentially the same as 
the Standard Metropolitan Area. ) 

Thus, we have attempted to develop an understanding of the subnational 
economy by focusing attention on various types of regional groupings, in each 
case seeking to employ the designation which is most revealing and most 
meaningful for the subject at hand. The groupings include the following: 

1. The great regions: the Northeast, the Southeast, and the West. 

2. Eight multi-state regions (particularly for the historical section). 

Tor a discussion of these and other problems of spatial and temporal units, see E. S. Lee, 
A. R. Miller, C. P. Brainerd and R. A. Easterlin, Population Redistribution and Economic 
Growth, United States, 1870-1950 (Philadelphia: American Philosophical Society, 1957), 
pp. 10-14. Other historical data considerations are discussed in the statistical appendix to 
this study. 



8 / Historical Review of Regional Economic Growth 

3. Modifications on the multi-state regions: as in reference to Upper New 
England or the North Plains or the Gulf Coast area. 

4. The individual states. 

5. State Economic Areas, made up of metropolitan and non-metropolitan 

SEA's. 



Time Segments 

Like a too rigid treatment of regions, a too rigid treatment of the time 
element can be misleading in economic description and analysis. There are 
important problems involved in the selection of the base year and in measuring 
values at two points in time. The length of the period of analysis is also 
significant in this regard. We might note as an example of this the significant 
difference that a single year can make with regard to the income picture. 
Between 1949 and 1950, state income increases ranged from a rise of 4% in 
Oklahoma to an increase of 23% in Montana. Thus, if Montana's income 
growth is assessed by comparing 1949 with, say 1940, an increase of 145% in 
total income payments will be recorded, but if 1950 is used, the increase over 
1940 will be shown as 199% — a difference of 54 percentage points. 

Since the problems referred to above are discussed in detail in the economic 
and statistical literature, we need not go into them here. However, we want 
to indicate that awareness of their importance has induced us to design our 
study so that subnational economic growth can be seen over various time 
segments, and with changes in base years and end points wherever appropriate 
for purposes of description and analysis. There are, of course, practical limits 
to the number of time variations that can usefully be employed in analysis, 
and — as in the case of regional designations — we have tried to strike a sensible 
balance. Thus, there are essentially two periods to which we have given our 
principal attention: the 80-year period between 1870 and 1950, which is the 
concern of Part III of the book, and the more recent period, 1939-54, which 
is employed for the analysis of locational shifts in Part IV. But we have also 
made use of other periods, either for purposes of testing or where the analysis 
could be enriched by doing so. 

We hope that we have at least avoided some of the more damaging intellectual 
traps inherent in the subject which is the focus of our study. 



2 / Economic growth since 1870 1 

{with a focus on the multi-state region) 



At the end of the Civil War the United States was still a partially developed 
nation on the fringe of the European-dominated world economy. By 1957, it 
had become a major force in international economics, holding a first-rank 
position among the nations of the world in the production both of raw materials 
and of manufactured goods. Meanwhile, in spite of recurrent war and reces- 
sion, its greatly enlarged population had come to enjoy the highest average 
level of living of any people in all history. 

However, growth did not occur uniformly throughout the country. From the 
beginning, the extension of population and economic activities into "newer" 
areas, as well as the relative transformation of "older" and more developed 
regions, proceeded at markedly divergent rates. Some of this differentiation was 
doubtless due to the local effect of business fluctuations or to the incidence of 
such factors as war and politics, but most of it stemmed from the spread of 
a highly productive population and of capital into areas of relatively unde- 
veloped natural resources in response to differential opportunities in various 
parts of the country. This territorial division of labor was undoubtedly one 
of the mainsprings of national economic growth. 

We want first to examine the migration of people and shifts in economic 
activities within the United States during the period from 1870 to 1957 and 
to trace concomitant changes in the regional structure of the national economy. 
Our purpose here is to provide a picture of how the American economy evolved 
into its present geographic pattern. The broad outline of this evolution may 
be represented by several historical series reflecting population and income 
changes among the regions over these years. These permit the tracing of the 
physical occupation and settlement of the continental territory, the relative 
growth in economic activities, and the relative changes in levels of living over 
these years. The analysis of the changes observed is reserved for later 
chapters. 

because of the lack of certain current data (i.e., post-1950 census), and for other reasons, 
the historical treatment of regional economic growth in Part III of this book covers the 
period 1870-1950. In this introductory section, however, certain of the more recent figures 
(for 1957) are included so that an up-to-date picture can be presented. 

/ 9 



10 / Historical Review of Regional Economic Growth 



Growth and Shift of Population, 1870-1957 

The movement of population and of economic activities in the United States 
since 1870 has been, in important part, a "spreading out" and "filling in," 
almost like the movement of flood waters over a rugged terrain. But it has not 
been simply a movement West, for there has been great and intensive develop- 
ment in the older settled sections of the East as well during these years. Thus, 
what we have is a nation still growing in all its parts. 

Tables 1 and 2 and Figures 2 and 3 show the absolute numbers and relative 
rates of growth in population, by regions, between 1870 and 1957. The 
absolute increases were heaviest in the Southeast, Great Lakes, and Middle 
Atlantic regions throughout the period. The rates of increase, however, were 
highest in the Mountain, Southwest, and Far West regions. The regions east 
of the Mississippi surpassed the national percentage increases in only a few 
decades, and then only at slightly above average rates. 



Millions 



m 



J> 






y 



> 
















Figure 2. Population at Census Dates, by Region, 1870-1950 (and 1957). 

Source: Table 1, note. Unless otherwise stated, sources for all charts are to 
be found in nearby tables. 



Economic Growth Since 1870 / 11 




Heavy lines indicate national 



Figure 3. Decennial Rates of Population Increase, 1870-1950 (and 1950-57). 



Before 1890 the Plains region also had a higher rate of population increase 
than the nation as a whole, but after 1890 it fell below the national average. 
In-migrants were fewer than out-migrants in Iowa and Missouri by the early 
twentieth century and in Minnesota, Nebraska, and Kansas by 1920. Since 
1910 the proportion of the national population residing in the Plains states 
has declined by 4 percentage points. In recent decades the "westward" move- 
ment of population has been mostly towards the Far West, Southwest, and 
southern Mountain states. 

Areas which consistently experienced the most rapid relative growth during 
these years were, on the whole, the least densely populated parts of the country. 
This is suggested by a comparison of the data in Table 2 with those in Table 3, 
which contrasts regional densities per square mile at various dates with the 
average density for the nation as a whole. As already noted, however, in 
absolute terms, the population increases were persistently heaviest in the Middle 
Atlantic, Great Lakes, and Southeast regions. In other words, the great bulk 
of the population continued to cluster in areas east of the Mississippi River. 
This can be seen in Figure 4, showing the regional distribution of the national 
population at various years since 1870, and in Figures 5 and 6 2 , showing the 
patterns of settlement in 1900 and 1950. 

The eastern part of the country — the New England, Middle Atlantic, Great 
Lakes, and Southeast regions — contained 85.5% of the nation's population 
in 1870, 74% in 1910, 71% in 1950, and 70% in 1957. In other words, the 
peopling of the West since 1870 has cost these four regions some 16 percentage 

^Figures 5 and 6, prepared by the Bureau of the Census on the basis of data for minor 
civil divisions, highlight among other things the importance of water supplies in the 
settlement of the arid West. 



12 / Historical Review of Regional Economic Growth 



tr- 
UO 

On 
r— i 

a 

ccj 

CO 



en 

C 

o 
'5b 



p 
o 
•-D 

Ph 

o 

Ph 



l> 


OS 


LO 




OS 




i-4 




6 




00 


© 

ON 


i— i 


as 



p 

§• 

Ph 

c/5 



c 
o 

•rH 

-Q 

T3 






§ 



o 



^ 



© 

©" 

o 



o 



o 



© 



o 
©' 

o 



$ 



LO 


CO 


ON 


rH 


3 


o 


CO 


On 


o 


00 


CO 


©^ 




CO 


CO 




VO 














LO 


r-~ 


LO 


CO 


<* 


00 


CO 


CO 




i— 1 




I— 1 



00 


© 


On 


«* 


o 


cm 


ON 


r^ 


cm' 


p' 



CO 


CM 


"tf 


CO 


N© 


CO 


cm" 

CM 


o 

CM 


cm' 

CM 


ON 


t>* 


cm' 



o 


CM 


On 


CO 


TJH 


CO 


LO 


CO* 


o" 


cm" 


©' 


t~-* 


CM 


t^ 


CM 


CM 


CM 











t-- 


CO 


VO 


CM 


00 


vd 


CO 

CM 


o 

CM 


CM 
CM 


o 



p 


r-l 


CO 


o 


ON 


O 


Tfl 


CO 


h- - 


CO 


©' 


CO 


r_H 


t^ 


CM 


LO 




CM 


CM 


CM 


1— 1 









O 


oo 


ON 


CO 
CM 


ON 


CO 
CM 



T? 


LO 


O 


i— 1 


NO 


LO 


r^ 


CM* 
CM 


CM 


LO 

CM 


CO 

I— 1 


LO 



LO 


Tf| 


LO 


CM 


rf 


CM 


r-H 


LO 


rf 


Tj* 


CM 


CM 


CM 


r-H 





O 


**. 


CO 


CO 


CO 


LO 


ON 


CO 


CO 


CO 


CM 


l>" 


cm' 


CO 


© 


CM* 




CM 


CM 


CM 


1—1 









CO 


c^ 


ON 


I— 1 


r^ 


NO 


Tfl 


CO 


oo' 


CM 


cm" 

CM 


On" 
CM 


ON 


CM 


© 


r-H* 


CO 
CO 


CO 

cS 
oC 


3 

r-H 

On" 


8 

5 


cS 

CO 


CM 


r- 1 


00 



csC 
3 

w 

2" 



P-5 03 

<5 erj 



I a 

° I 

co Ph 



© 


1 


3 

o 


-3 
03 


3 

o. 




co 






ta 






o 


O 

3 


3 
o 


rH 

CD 

s 

3 


CO 
O 




Oh 


erj 
03 

H 

3 


ft 

c0 


3 

CD 


""3 

03 
co 






CO 


3 


O 


"o 




V 




•2 


CD 


03 




03 
O 
U 


P 


ta 

3 


CO 

-3 

CO 


rH 

ft 

CD 




3 
O 


03 
03 

co 


O 

3 




rH 

o 




3 




_C 


co 


CO 




en 












3 

03 
O 


03 
"5 


-3 

-3 


h3 
o 
c0 


rH 








CO 


CD 






3 

s 


CO 

C 
in 
CD 


CO 

H 


o 


CD 

3 

r3 


CO 


"C 


-3 




3 


"c0 




o 








> 


ft 


03 

"3 


3 
O 
CD 


P 


O 


c0 


CO' 


£ 


CO 


03 

O 




O 
3 


r^ 

3 


o 

■Jh 


3 


^ 


3 
o 


s 

CD 


c0 
f-H 


03 




cm" 


"5b 

03 


£ 


CO 

!5 


cO 


CO 






2 


On 




03 










-o 

03 


£ 


_03 


CO 

03 


cO 

-o 


^ 


to 

3 

03 


bJ3 

CD 


03 

CO 


| 

3 


4-> 

3 


CO 

3 
fcJJ 

3 


03 

3 


13 


Ov 


O 
CO 


3 


< 


QD 


3 

rH 


ON 


03 
-3 


.2 
to 




NO 


J^» 


d 


# 3 




<o 


03 


05 


Ph 


CO 


13 


0Q 


3 


£ 


o 


c0 


•4-1 




co 






03 


o 


H 




c 


H 




cu 




CO 


o 




"3 


fc. 




l> 




-3 


C 


S 


4-1 

CO 

Ph 


o 
o" 

NO 

CM, 

<H-I 

o 


3 
h3 

CO 

CO 


3 
c0 

3 

.2 
'5b 

03 


^3 
03 

s 

03 

-3 

2 


© 


O 


3 


»o 




o 


CO 


> 

o" 

LO 
ON 

I— 1 


Sh 
CD 

s 


CO 


cS 
03 

,3 

3 

o 


cO 

03 


03" 
O 

rH 
03 

s 


a 


3 
3 

03 
03 


I 


CO 

u 

3 
o 

«-M 

o 


-3 


B 

o 


ea 
3 


na 


co 


cO 
co 


3 


o 


. 


-S3 


CO 




3 


O 


O 

03 

-73 




03 

e3 

"eft 


eft 

3 
H 


03 

s 

CO 


CO 

3 
co 
3 

CD 


c0 

s 

03 


CD 


CM 

03 


s' 
o 

4-> 


ft 
CD 

Q 


CJ 


fH 

03 


O 


r3 


43 


CD 


to' 


IS 


eo 


b» 


3 
ft 


P 


P 

-3 
■i-« 
l> 

H 

a 


CO 

3 

03 
g 

CO 

.3 

*5* 
co 


_CJ 

to 

"JS 

e 

CO 

1 


3 
O 

s 

CO 

3 


o 

ft 

CO 

CO 

Cfi 

pa 


S 
03 
LH 


-73 

CD 


M 

CO 
03 


o 
•2 


CO 

-3 


3< On 


ft 

cO 


,>* 


CE3 


3 


03 
CO 


O 

«+H 


-3 


ea 




03 


s 


CO 


< 


r3 


3 


03 

rJ3 


03 

^3 


2 




rH 


CO 

3 


CO 


3 


& 


W 


o 


03 


CO 


o 






HH 


u 


-3 


CD 


s 


& 












o 












CD 













Economic Growth Since 1870 / 13 



6 

On 



c 
(0 

o 
On 

r-H 

© 

CO 



>% 

^Q 



o 



§ 

c 
.2 

O 



"3 

CD 

8 



e2 



o 




g 


©' 


o 




o 




CO 




1—1 




© 




Os 


LO 




lO 


CO 


<N 



in ©' 



"* 


© 


CSI 


© 


© 


°Q 


CO 


CO 


© 


rH 


"<* 


<M 


<sf 


NO 


•n 


co 


i-H 


l— < 


i— 1 






"H 


•*H 


^ 



CO 


CO 


CO 


n© 


r- 


t^ 


en 


en 


m 


m 


<o 


r-i 


t^ 


On 



O 


CO 


t^ 


in 


00 


"*■ 


On 


*«H 


CS 


»n 


t>* 


© 


t^ 


*"H 


NO 


<N 


1—1 


*-H 


"*H 


i— i 




CM 


<N 


«o 



<N 


VO 


<N 


NO 


m 


"* 


o~ 


co 


h- 


CO 


t* 


m 


<N 


§ 


CO 


en 


l ~" 1 


Cn 


i— l 






»n 


t^ 



© On NO CO 
On CO 00 CO 

—* CM f-> i-H 



CO N© "-H C2> 

1/5 N «5 N 

i— i uo rj< cm 



CO 


CO 


CO 


•-H 


NO 


-~H 


LO 


© 


© 


c- 


in 


NO 


CO 


^ 


cm 


CM 


I— 1 


^ 


*n 




NO 



© 


^ 


CO 


CO 


t^ 


NH 


m 


ON 


CM 


r^ 


On 


CO 




i-H 


CM 


i-H 


in 


t^ 



c 



CD 






'3 


-C 




J3 


q 


3 


'S3 


3 


p 


O 
C/3 


Ph 


O 


£ 



o « 

■a § 



.3 H 

•« CD 



14 / Historical Review of Regional Economic Growth 

Table 3. Density of population, by region, 1870-1957 

(Population per square mile of land area) 



Region 


Area* 


1870 


1890 


1910 


1930 


1950 


1957 


United States 


2,974,726 
















sq. mi. 


13.4 


21.2 


30.9 


41.3 


50.7 


57.3 


New England 


(2.12) 


55.2 


74.4 


103.7 


129.3 


147.5 


156.3 


Middle Atlantic 


(3.78) 


87.6 


125.8 


188.1 


254.5 


299.1 


326.6 


Great Lakes 


(8.23) 


37.3 


55.0 


74.5 


103.3 


124.1 


143.1 


Southeast 


(17.95) 


21.7 


30.1 


41.2 


51.1 


63.3 


69.7 


Plains 


(17.17) 


7.6 


17.5 


22.8 


26.0 


27.5 


30.0 


Southwest 


(19.08). 


1.8 


4.8 


10.7 


16.0 


20.0 


23.6 


Mountain 


(17.23) 


0.3 


1.8 


3.9 


5.4 


6.8 


8.1 


Far West 


(14.44) 


1.7 


4.5 


9.9 


19.3 


34.1 


43.5 



* Figures in parentheses show proportion of total U. S. land area 1950. Figure in italics 
are above the national average. 

Source: Statistical Abstract of the United States, 1954, Table 4, p. 10; also see 
Table 1, note. 



points of their share of the nation's total population, and since 1910 — it is 
particularly worth noting — only about 4 percentage points. The Great Lakes 
region, in fact, has a larger proportion of the nation's population at the present 
time than it had in 1910. 

The region which secured the greatest share of the relative population 
losses of the East and the Plains region was the Far West. Its share of the 
national population increased from 4.7% to some 11% of the total between 
1910 and 1957. 

If regional growth is measured in terms of employed labor force rather 
than in terms of population, almost exactly the same pattern emerges — not 
unexpectedly. That is, between 1870 and 1957, the greatest relative expansion 
took place in the Far West, Southwest, and Mountain regions: the least, in the 
Southeast and New England. 

The shares of the national labor force compare quite closely with the popu- 
lation shares for all regions (the greatest deviation being in the Southeast, 
where the influence of high birth rates emerges as significant). This can be 
seen by comparing the population and labor force shares of the Middle Atlantic, 
Far West, and Southeast regions for 1870 and 1950: 



Economic Growth Since 1870 / 15 




O oo O o r^ 
r- & - o in in 
co cc 01 o> 01 oi 



Figure 4. Regional Distribution of Total Population, 1870-1950 (and 1957 
Source: Table 1, note. 



Share of U. S. Total 
Middle Atlantic: 
Population 
Labor force 
Southeast: 
Population 
Labor force 
Far West: 
Population 
Labor force 



1870 

24.7% 
25.2 

29.1 
28.0 

1.8 
2.5 



1950 

22.3% 
23.5 

22.4 
20.7 

9.7 
10.1 



Urban Growth 

Throughout the period under review there has been a continuous decline in 
the proportion of total population living in rural areas and a marked increase 
in urban population. The rate of urban increase has been positive in every 
region at every census since 1870. Urbanization in all parts of the country 
has been a central facet of recent American history. This significant change has 
accompanied the industrialization of the nation. 

The decline in the share of the labor force engaged in resource activities 
(agriculture, forestry, mining, fishing) is traced in later chapters, as are the 
great increases in manufacturing and service employment. At this point we 
shall note only the profound effect that the changing structure of economic 
activities has had on the urban-rural pattern of population and in turn on the 
regional pattern of settlement. These are recorded here because of their 
significance for the analysis of regional economic growth which follows. 






Economic Growth Since 1870 / 17 

The increase in urban population within the various regions between 1870 
and 1950 is summarized in Tables 4, 5, and 6. 3 The relative size of each region's 
urban population at various dates is shown in Table 4. By 1950, it can be 
seen, six of the eight regions had a larger urban than rural population. The 
decennial increases by regions are shown in Table 5. The rates of urbaniza- 
tion were higher in the "newer" and less developed regions, for the most part, 
than in the older and more developed areas of the country, though we tend to 
think of newly settled territory as largely rural. 4 Table 6 permits a comparison 
between the rates of urbanization and the proportion of urban residents in the 
total population of each of the regions. In spite of the very rapid increase in 
urban population in the West, the great bulk of the urban population resides 
in the East, and particularly in the Middle Atlantic and Great Lakes regions. 

Viewed from the standpoint of city size (Table 7), the general process of 
urbanization has assumed different patterns in different regions. In 1870 only 
the Middle Atlantic and New England regions had the clearly denned begin- 
nings of an "urban hierarchy" — that is, a broad base of many smaller-size urban 
centers underlying a smaller number of larger cities ranging up to a dominant 
regional metropolis. An emergent hierarchy was also evident in the Great Lakes 
region, but, in view of the much greater land area involved, these four or five 
midwestern centers with a population of over 50,000 could better be regarded 
as sub-regional centers in their own right with no one of them yet dominant 
over the others, though Chicago was already moving into that position. The 
same was broadly true of the urban scene in the South. Elsewhere in the 
country, the number of cities with more than 50,000 inhabitants could be 
counted on the fingers of one hand. In 1870, the Southwest and Mountain 
regions still had no center above 15,000 and the Plains had only three — St. 
Louis, St. Paul and Kansas City. In the Far West, San Francisco stood alone, 
rather like one of the great eastern coastal cities of colonial days. 

By 1910 the force of urbanization was being felt throughout the entire 
country. The New England, Middle Atlantic, and Great Lakes regions had 
developed well-established urban hierarchies. The rest of the regions, with 
agriculture, forestry, and mining much more dominant than in the northeastern 

3 By 1950, the older census classification for presenting data on urban communities tended 
to understate the size of urban population, and a new classification in terms of "standard 
metropolitan areas" was introduced. In the present study, however, we have adhered to the 
older designation since it facilitated comparisons over time. For historical data based on 
the new designation, see Amos H. Hawley, The Changing Shape of Metropolitan America: 
Deconcentration Since 1920 (Glencoe, 111.: The Free Press, 1956). 

4 It is interesting to note that urbanization has proceeded in spurts, and has been cyclically 
oriented. Outstanding are the great spurts of the Mountain states during 1870-90, the Far 
West in 1870-90 and 1900-10, the Plains in 1880-90, and the Southwest over the four 
decades from 1870 to 1910. 



Figure 5 — opposite. Distribution of the Population, 1900. 

Source: U. S. Bureau of the Census, 1900 Census of Population, Statistical 
Atlas, plate 13. 



Economic Growth Since 1870 / 19 

Table 4 Urban population as percentage of total population, by regions, 
1870, 1910, and 1950 



Region 


1870 


1910 


1950 




% 


% 


% 


United States 


25.2 


45.7 


59.0 


New England 


44.4 


73.3 


74.3 


Middle Atlantic 


44.1 


70.2 


74.0 


Great Lakes 


21.6 


52.7 


65.7 


Southeast 


9.5 


19.4 


42.5 


Plains 


18.9 


33.2 


49.9 


Southwest 


6.9 


22.5 


55.5 


Mountain 


13.9 


40.7 


51.8 


Far West 


31.2 


56.0 


62.7 



Figures in italics are above the national average. 

Source: Adapted from 17th U. S. Census of Population, 1950, Vol. II, Pt. 1, Table 
15. The additions to total population for the Southeastern states and Texas (see 
note to Table 1) for 1870 have been allocated as urban or rural according to the 
weight of the urban population shown in the original census report. This tends 
to exaggerate the absolute size of the urban population in some Southeastern 
states and in Texas but preserves the relative proportion in each case. The urban 
proportion in 1950 follows the standard census definition of "urban" (comparable 
with 1940) rather than the new definition adopted in the 17th U. S. Census, 1950. 
See also Tables P-4B and P-4C of E. S. Lee, A. R. Miller, C. P. Brainerd, and 
R. A. Easterlin, Population Redistribution and Economic Growth, United States, 
1870-1950 (Philadelphia: American Philosophical Society, 1957). 



part of the nation and with huge land areas involved, were developing a dif- 
ferent pattern of urban centers. There were several rapidly growing centers in 
each of these regions, but they were far smaller than the metropolitan com- 
munities of the Northeast and were not part of a clear-cut urban hierarchy. 
In the Southeast, for example, only New Orleans had a population of over 
250,000 by 1910. 



Figure 6 — opposite. Distribution of the Population, 1950. 

Source: Conrad and Irene Taeuber, The Changing Population of the 
United States (New York: John Wiley and Sons. 1958), Figure 3c. 



20 / Historical Review of Regional Economic Growth 



3 

a 

a 



a 
3 

Q 



to 



o 


f- ( 


in 




o\ 




i— i 




6 




i> 




CO 


O 


r-4 






8 


O 


r— 1 


*i— i 




bD 




V 




M 






o 

o 
os 



8 



c 

93 



S 



LO *0 pH 

2 8f 5 



*0 


«> 


v© 


cm' 

»o 


p>h 
«9 


«T5 
*<* 



On co 



CO t~- t^ ^O SO co 
CM © CO On t^ ^0 

<-> cm cm ?o f- 1 *o 






© 


t^ 


t^ 


CO 


CM 


v© 


CO 


»*H 


t^ 


cm 

cm 


8 


«o 


cm 

CM 


«5 

SO 


CO 
CM 


58 



cm 


<o 


CO 


*o 


NO 


•—* 


CO 


CO 


l-H 


CO* 


CO 


cm' 


CO 


^J- 


co 


"H 


so 


N 



in 


o 


cs 


»*> 1 


t^ 


On 


SO 


^ 


co 


in 

CO 




is! 

«0 


si 


to 

VO 


lO 


n! 



00 SO 

co On 

r* CO 



C£> CO CM CM 

so co sp io 
so so cs co 



ON •— t 
On On 



r-; 


r^ 


CM 


CM 


"•«< 


VO 


*o 


P-H 


to 

CO 


"#' 

co 


SO 
lO 


sd 

i— 1 


fO 
VO 


fo" 

PO 

•— 1 


*•* 
fo 


On 
CO 



-o 


fi 










c 


CO 


CO 








cd 

Ho 




0) 










ed 


4-1 

CO 




CO 


c 




J 


CtJ 




CD 
pg 

"3 




4) 

-o 




pg 

"3 


in 

.s 

°cd 


85 


s 




o 

CO 


E 


o 
C/} 



.3 «o 



cd rj< 

co co 

cd ,ed 

.■a h 

.a <u 

£ 

3 g 

bC £ 

fa D 

W O 

co 



Economic Growth Since 1870 / 21 



a 
o 

i 

I 

o 

Oh 



■a 



(fi 



en 
§ 

"■B 

o 
o 

u 



1 

o 



o 

c 

43 



43 

en 

C 

o 
'So 



u 

"S 

o 



o 

. lO 
rH 

e co 

E-H I— I 



J5 3 

rt o 





CO 


VO 


irt 


SS 


s 


5? 


8 








s« 








§*> 










^ 


£ 


w 




c/3 


t*H 


t/i 





fO 


O 


fl 


irt 


CN 


o 


co 


<N 








2 S 
§ S. 







£ 1 

2 ft 
2 

PL, 



CO 0\ 00 rH © 

e£ *•» i"* ' t* c>i 



^ ^ « * ~ 

fe fti iz; co* S 



O i-h lO O CO 



J "< w 

3 £ d S 85 



<N «N t^ CO O 

o o\ to co oi 



r£ 



^ 



Oh fe Cfl S 





CO 


1— t 


vO 




CO 


<N 


CO 


o B 




^ 


VO 


rH nj 








§ tf 








2^ 










fc* 


£ 


«J 




CD 


£* 


s 



(N CO CO 

CO rH LO 

co co CN 



o 
o 





to 


LO 


I— 1 


CN 


rH 


t^- 


t^ 


CN 


o « 
CO ca 


LO 

CO 


3j 


On 

CO 


LO 


co' 

LO 


LO* 

CO 


Tj* 

CO 


o 


O pi 
o 

a* 




^' 


^ 


J 




w 


< 


w" 




si 


c/5 


ps* 


d 


s 


z 


£ 


co 




5 


£j 


CO 


T? - 


LO 


s 


r>- 


00 



22 / Historical Review of Regional Economic Growth 



.5 



o 
to 
o\ 
i— i 

a 



o 

On 
i— I 

<D 
i> 

CO 



en 






+ 
000'000'T 



000'000'T 

ooo'oos 

ooo'oos 
-ooo'oss 

000'OSZ 
-OOO'OOT 

000'OOT 
-OOO'OS 



+ 

000'000'T 
000'000'T 

ooo'oos 

ooo'oos 
-ooo'oss 

OQO'OSS 
-OOO'OOT 

OOO'OOT 
-OOO'OS 



+ 

000'000'T 

000'000'T 
-OOO'OOS 

ooo'oos 

-OOO'OSS 

OOO'OSS 
-OOO'OOT 

OOO'OOT 
-OOO'OS 



0<M<N©00©i— I 



i-H T* CO r-H <N 



OCO'*'* i ^i <:, 0'-- |l O 



I— l CO O LO ^ ^© <— ILO 



©CM<-H©0000 



i-H(M^O>— IOOO 



©T?COl— Ir-HOOCSl 



r-^LOtovocooi— <r? 



<N 00 >— I V© lO lO >— ' r ~ ' 



o o o o o o 



o <— IOOOOOO 



,-Hi— i r- lO>— IOOO 



O^"— ifMOOO-— I 



CSICOcO'— fOOOO 



£; g O co H, to - S fc 



CO 


00 




GSJ 


© 


o 


00 




1 — 1 


d 




p. 


w 


o* 


OS 


>-"• 


& 




o 




CO 





Economic Growth Since 1870 / 23 

By 1950 manufacturing activities had become significant in the Far West 
(particularly the Coast area) and in parts of the Southeast, and both of these 
regions had begun to develop more sharply defined urban "systems/' In the 
rest of the country, except for eastern Texas, the larger cities tended to be 
more isolated regional centers (in some cases resembling "oasis cities"), with 
hardly a discernible hierarchy of smaller communities. Thus the population 
map of 1950 (see Figure 6) reveals three great urban masses highlighting the 
three great manufacturing zones of the nation — the major Manufacturing Belt of 
the Northeast and the lesser manufacturing zones of the Pacific Coast and 
the Southeast. 



Changes in Total Income, 1880-1957 5 

One of the most comprehensive and suggestive measures of regional 
economic growth is provided by total personal income. 6 Table 8 presents data 
on changes in the regional distribution of the nation's total personal income 
between 1880 and 1957. In some important respects these data show an 
over-all pattern of differential regional growth similar to that shown by the 
data on population in Table 1. Like the changes in population, the changes in 
income reflect the impressive growth in volume of economic activities in the 
Far West and Southwest regions, the gradual decline in the volume shares of 
the New England and Middle Atlantic regions, and the stability of the Great 
Lakes region over the period as a whole. 

A closer look at the data reveals, however, that the distribution of total 
income among the various regions does not correspond fully with the popula- 
tion distribution. This is particularly true for the Southeast. In 1880 this 
region had 27% of the nation's population but less than 14% of the total 
national income; and in 1957, while it had almost 22% of the total popula- 
tion, it had only some 15% of the total income. The differing pattern of 
regional shares reflects the sizeable differences in income-earning capacity of 
the activities dominant in the various regions. This was particularly important 
in the earlier part of the period. 7 

'Unfortunately, for the earlier part of the period under examination, the years for which 
income data are available do not coincide with the years for the population series discussed 
in the preceding sections. The best figures on state income available, those devised by 
Richard Easterlin of the University of Pennsylvania, begin with 1880 and proceed by two- 
decade intervals to 1920. Thereafter, income figures are available by decades in the Com- 
merce series and after 1929, in fact, on an annual basis. 

"Ideally, to measure growth in the volume of regional economic activities, we would want 
annual data on production — on net value produced- — within each of the regions. But we do 
not have data on gross or net product for any subnational units. However, the Department 
of Commerce series on state personal income does provide a fairly close approximation to 
the desired measure. 

7 Cyclical movements, especially in agricultural prices, have also played a significant role 
in the levels of total income within the various regions. 



24 / Historical Review of Regional Economic Growth 



o 

lO 
r— I 

§ 

CO 



S3 

o 

• — t 

r9 
***"! 

to 

"a 
o 
"5b 






s 





VO 




vo 


lo o 




T? o 




co i— ) 





LO 


LO 


rj< 


»—l 


CO 


(N 


O 


ITS 

cm 


cm' 

CM 


LO* 


CO 


VO 


CM 


CO 
r-t 





h~ 


<* 


in 


CM 


CO 


LO 


CM 


r^ 


LO o 
CS| o 

CM l-H 


vd 


cm 


CM 
CM 


LO 


CO 


*d 


cm' 


r-5 

i— i 





CM 

CO 


LO 


r- 


CM 


CO 


CM 


o 


o\ 


CM K o 
CM £^ 


d 

CO 


S3 


CO 

l-H 


cd 


to* 


cm' 


o" 


LO O 


















CO o 


















<^ w 



















VO 
VO O 



CO 


VO 


CM 


Ov 


CO 


o\ 


CO 


CO 


CM* 


r-4 


CO 


Tf 


i-5 


cd 


CO 


CM 


i -1 













CO 


CM 


CM 


O 


CO 


r- 


LO 


VO x-v 


CO 


d 


CM 


CO 


O* 


LO* 


CM 


CM ©^ 




CO 


CM 




i— 1 




















VO <=>. 





















On 


CO 


rf 


O 


CO 


CO 


LO 


CO 


r-H 


'"-* 


Ov 


© 


CM* 


CM 


CO 


CO* 


CM 


LO* 


CO 


o 




CO 


CM 




l—l 








LO" 


o 



















CO ^o i— i CM 

Tji ^. l-H CO 

CO o 



CO 


t^; 


i-J 


l-H 


** 


o 


CM* 


CO* 


_4 


cm" 


H 


LO* 


CM 




i— 1 









0) 

CO 


-d 


C 






c 

cd 
"hfi 


at 


cS 


CO 




a 




hJ 


cS 




is 


"■3 
-d 




3 


1 


0> 


s 




o 

co 



I I I 

Ph CO S 



^ 



"g 



* o> 



»Q l—l 




V 4) 




S3 




© 0) 








*-« 








»»» /— V 




e vo 




e lo 




O Q\ 




52 -• 




V 




ft, of 




o 




•h'S 




^ O 




« .5 




-£3 +3 




ca fl 




fH T 1 




o£ 




W ti 




a 




• a> 




« S 




m S 




a » 












03 .. 




is a 




^ 2 




*.» 








**:1 




°t 












fo „ 




LO J> 




f- s 




. 52 




ft I 




. s 




rH o 




J^CJ 




93 MH 




i— i O 




,o 




S3 +-> 




h g 




^ a 




t? ^ 




03 




03 Q 


CO* 

1— ( 


H . 




H CO 


d 


o 




2 


r- r 

03 


2 » 


3 
C3 


43 g 


H 


O 


00 


iw 


ON 



w 



m 3 



PQ 



O 05 

a"! 

ON {». 

a a 

co vVj 



Economic Growth Since 1870 / 25 



Changes in Per Capita Income 



The interplay between population growth and growth in total income within 
the various regions is reflected in the per capita income levels and rates of 
increase. Table 9 shows estimates of real income per capita (1929 dollars) 
over the 1880-1957 period in each of the regions. There has been a striking 
trend toward equalization among the regions. In 1880, regional averages ranged 
from 211% of the national average to 50%. In 1957 the highest regional 
income was only 119% of the national average and the lowest 70%. 

Since 1920 the regions in which manufacturing and service activities are 
predominant have had considerably higher incomes than the less industrialized 
Mountain, Plains, Southwest, and Southeast regions. Before 1920 the advantage 
of the more urban-industrialized regions was by no means so consistent in 
this respect. In the last two decades of the nineteenth century the Far West 
and Mountain regions had the highest per capita real incomes in the country; 
it was only after 1900 that the industrial areas markedly improved their 
relative position. First the Middle Atlantic and New England regions, followed 
after World War I by the Great Lakes region, closed the gap between their 
income levels and those of the Far West. Between 1920 and 1940, the Middle 
Atlantic region was actually the pacemaker for the nation. 

More recently, the states and regions appear to be grouping themselves into 
three categories with regard to per capita income levels (Figure 7). These 
groupings are: (1) the urban-industrialized-service regions of the Manufactur- 
ing Belt and the Far West, with per capita incomes above the national average, 
and mostly well above the average; (2) the Mountain, Plains, and Southwest 







High income stales 
(above national average) 

£23 Med 



jium income stotes 



(below 1950 national average or $1491) 



Figure 7. Per Capita Personal Income, 1957. 



26 / Historical Review of Regional Economic Growth 

regions, northern New England and the peripheral Southeastern states of 
Florida and Virginia, with incomes somewhat below the national average; and 
(3) the greatest part of the Southeast, standing alone (except for the Dakotas) 
in terms of extremely low per capita income levels. 8 

The groupings are fairly well defined. Within the high-income areas, only 
three states in the Northeast (Indiana, Wisconsin, and Rhode Island) and only 
one state in the Far West (Oregon) were below the national average in 1957. 
In all the other states of the Northeast and the Far West, per capita income 
ranged from $2,112 to $2,821. The range in the low-income states of the 
Southeast was from under $1,000 (Mississippi) to slightly over $1,400 
(Georgia) ; the average for this region in 1957 was below the national average 
of seven years before ($1,451). The middle-range group covered the relatively 
narrow range between $1,531 (South Dakota) and $1,996 (Colorado) in 
1957. 9 

The depressing effect of population growth keeping close pace with the 
growth in total income is evident in the Southeast where, despite steadily 
increasing economic activity and increasing wage levels, per capita income 
continues to be far below the national average. In contrast, the slackening of 
population growth in the Plains states has helped to keep the level of real 
income per capita from falling sharply in relation to the national average in 
spite of a relatively slow growth in the volume of economic activities and total 
income. 



8 A few of the states are at the periphery of these major classes; these are shown on 
Figure 7 by dots in the case of the lowest income class and a break in the solid gray in the 
case of the states slightly under the level set for the highest income class (i.e., the national 
average per capita income in 1957). 

9 State per capita income figures for any one year inevitably reflect special circumstances, 
so that some variations are to be found from year to year. This is particularly true because 
of the wide cyclical swings in farm prices and farm incomes. For this reason, the states that 
are on the borderline between broad income categories are specially marked in Figure 7. 
For example, Wyoming, which was above the national average in 1957, was below the 
average in 1955 and 1956, and therefore is marked as borderline. This is true also of 
Indiana and Rhode Island which were just below the national average in 1957, but above the 
average in every other year between 1950 and 1956. Oregon was above the national average 
during the period 1950-53; just short of the average during 1954-57. The states of North 
Dakota, South Dakota, Louisiana, and West Virginia are borderline between the lowest and 
middle income categories if the standard of catching up with the national average of some 
years back is employed. Of these states, all but North Dakota were just above the 1950 
national-average level (of $1,491) in 1957, but below this figure in 1955 and 1956. Of the 
Southeastern states, only Virginia and Florida can be said to fall clearly into the middle- 
income category. 



Economic Growth Since 1870 / 27 



© 




-o 






r^ 


o 


LO 


cm 


On 


On 




„ 




CO 












ft 




« 




O 




J-l 






O 

LO 




On 


CJ 




s 




O 





b£ 



WD 
M 

03 w 



-a 

c 
cd 

t> 

LO 



c s 



a 

o 

CO 



On 





tf. 


CO* 


CO 


CM 

ON 

o 


© 


ON 

ON 

00 


CO- 


© 

ON 


co 


o 




CM 


t^ 


CO 


r- 1 


vo 


CO 


CM 


t^ 


NO 

CO 
CSI 




s 


LO 


CO 

LO 

<o 


O 

co 


s 


CO 

s 


CO 


vo 

x* 



c£ 



CM 

On' 
O 


oo" 


**. 


VO 


On 


CO 


00 

LO 

On 


© 

CM 


** 


CM 


On 


CN 


VO 


VO 


OO 


^ 


C5 

vo 
►—I 


>o 

CN 


CO 
CO 
l—H 


On 


CO 
O 
O 


On 


o 


On 
CM 





*■ 


CM 
l> 
CM 


CO 

cm' 

CO 


cm' 

r— 1 


v© 
r-' 

LO 


CM 

00 


CM 

©" 


CM 

ON 

00 


ON 

l-H 

CO 


x* 




x* 


CO 


x* 


CO 


LO 


o 


CM 


t^ 


l> 
CM 




LO 

CM 

On 


LO 

NO 

On 


LO 
CO 


On 


8 

LO 


LO 


9 

VO 


On 
LO 

On 





* 


CM 

On 
CM 


vo 

CM 


vO 

© 


CM 

O 

LO 


00 


CO 
<* 

vo 


CM? 

vd 
CO 


CO 

§ 


O 




ts. 


On 


VO, 


o 


i— i 


CO 


00 


CO 


o 




CM 

co 


CM 

On 




CM 

co 


CO 
CM 
LO 


i-H 
<* 


r-H 
LO 
LO 


CO 



c£ 



u 



•<* 

CM 


CM 

CO 

CO 


o 


vo' 

LO 


oo 




CM 

cm' 

o 

i-H 


LO 

CO 
1— 1 


^O 


LO 


LO 


CM 


o 


CM 


l^ 


vo 


On 

—1 




CM 

vo 


VO 
CM 
CO 


© 

LO 


1 


LO 


1 



CO 


c^ 


<N 


CO 

CO 


CO 


s 


zn 


l-H 


l-H 


t^ 


CM 


VO 


r^ 


CO 


On 


LO 


t^ 


CO 


LO 


LO 


** 



00 


© 
t^' 


00 


CM 

4 


On 
CM 

vo 


X* 


On 


vo 
















t^ 


LO 


CM 


CO 


VO 


ON 


l-H 


s 


2 


s 


r-i 


CM 


vo 


VO 





O 

o 


CO 

© 


On 


CO 
CO 


LO 

8 


o 
vd 
vo 


CO 
CM 


co 


CO 


** 


VO 


CM 


f~ 


LO 


CM 


LO 


Z? 

CM 


CO 

CO 


o 

LO 

l-H 


CM 


CM 

00 


s 

LO 


CO 
CO 
VO 



C 


-d 


O 




bfi 




CO 


c 


Oh 


P 



-o 


c 












CO 


cfi 








JO 


< 


03 

-a 


tfi 




tn 


c 




^ 




<v 


w 


CO 




<V 




* 




H3 




J=S 


a 


A 


£ 


-o 


<D 


3 


3 


Z 


£ 


o 


O 
CD 


S 


o 

CD 



* 



CO CO 

00 x? 
l-H 

coco 



300 

"^ l-H f-H 



*3 CM ^ °^ 
2 On l> 00 

.Th rH LO tJ< 

s w J. 1 

"H » o o 
b99 c? 



5 00 



00 
cu 

CO 



3 © 



co eo 
10 ■ eg 
On * 



B On 



CO 



J? 8 

*£ o 



en .. 

'O <u 



2 Oh 5 
w ^ 

•8-1 > 



O h- 

CM LO 

On On 

r-i i-H 

cico 

00 CO 

00 On 



CO 

to CM* 

-° JJJ 

<V CO 

CD ^ 







Economic growth, 1939-1954 
{with a focus on the states) 



Much of our analysis of the factors behind regional economic growth will 
concentrate on a recent period of relatively short duration — the 15-year period 
1939-1954. 1 For this period we have examined in some detail the major 
elements suggesting changes in volume of economic activities within the states — 
population and total personal income — and changes in living levels as repre- 
sented by per capita income. State data have been used in this examination 
to give a more detailed and particularized view than the regional data alone 
provide. 



Measurement of Relative Change 

A whole series of data on relative (percentage- wise) changes in population, 
total income, and per capita income are brought together in Figure 8 for the 
states and in Figure 9 for the multi-state regions in order to show the interplay 
of the key elements in regional growth. This technique of presentation, based 
on the "relative growth" chart developed by Edgar M. Hoover and Joseph L. 
Fisher, 2 has the important advantage of highlighting the wide variety of growth 

^his period seemed to be appropriate for several reasons: (1) it coincides with two 
census dates for which considerable data are available of the type needed for the detailed 
analysis involved. (2) The years 1939 and 1954 have fairly comparable business-cycle 
(particularly unemployment) characteristics — a critical element in measurement and analysis 
of economic growth, although comparability of this type at best is highly tenuous. (3) 
Both the early part and the latter part of the period were influenced by war; first, World 
War II, and then the Korean War. Actually, of course, no years during the 20th century 
can be chosen (for growth comparisons) which can be considered to be really removed from 
the "unusual" factors associated with war and depression. 

2 Edgar M. Hoover, Jr., and Joseph L. Fisher, "Research in Regional Economic Growth," 
Problems in the Study of Economic Growth, Universities-National Bureau Committee on 
Economic Research (New York: The National Bureau of Economic Research, 1949), pp. 
195-203. 

In the two charts that we present here the solid horizontal axis measures 1954 population 
as a percentage of 1939 population and the solid vertical axis measures 1954 total personal 

28 / 



Economic Growth, 1939-1954 / 29 

patterns within the national economy. The various combinations of increases 
or decreases relative to the national average are seen to be as follows: 

I. Above-average increases in population, total income, and per capita 
income: The Southwestern states fall mainly into this category, with Arizona 
leading in both relative population increase and relative increase in total income. 
Florida follows closely on both these scores. In terms of relative increases in 
per capita income, however, the only states in this group that rank very high 
are New Mexico (seventh highest in the nation) and Texas (eleventh in rank). 

II. Above-average increases in total and per capita income; below-average 
increases in population: Most of the Southeastern states fall into this group, 
and the Plains states are also strongly represented (Kansas, Nebraska, South 
Dakota). In two states — Mississippi and Oklahoma — population actually 
declined between 1939 and 1954, while total income and per capita income 
increased more rapidly than for the nation as a whole. Oklahoma had the 
third highest percentage increase in per capita income in the country and 
Mississippi the fifth highest. 

III. Above-average increases in per capita income; below-average increases 
in population and total income: With the exception of Wisconsin, all of the 
states in this category are west of the Mississippi. The Plains states and the 
Mountain states (except Wyoming) that are not in either of the first two 
categories are in this one. When an average for the multi-state regions is 
struck, the Plains region is shown to be well below average in population 
growth, almost exactly average in growth of total income, and above average 
in per capita income increase, while the Mountain region had slightly above- 
average increases in all three over the 15-year period. 

income in current prices as a percentage of 1939 total income. The data are plotted on a 
double logarithmic scale. The origin of the diagram (A) is at 100 for population, which 
represents population unchanged between the two dates, and at 207 for total personal income 
at current prices. Since prices increased 107% between the two dates, as measured by the 
Gross National Product deflator, an increase in total money income of 107% between 1939 
and 1954 is the same as an unchanged total real income. 

Each state (or region) is represented by a point on the graph with coordinates deter- 
mined by its percentage changes in population and in total personal income between the 
two dates. Any point to the right of the solid vertical axis represents an increase in 
population and any point above the solid horizontal axis an increase in real total personal 
income. 

The solid diagonal line drawn through the origin is the locus of all points on the chart 
for which per capita income in current dollars increased 107%, or for which per capita real 
income remained unchanged between 1939 and 1954. Any point above this line represents 
an increase in per capita real income. 

In addition to the solid coordinate axes, dashed axes are shown. The origin of these (B) 
is the U.S. average increase — 23% for population, 291% for total personal income in current 
prices, and 217% for per capita income in current prices. (In "real" terms, the national 
average increases were about 89% for total personal income and 53% for per capita 
income.) Any point to the right of the dashed vertical axis represents a population increase 
greater than the national average, and any point above the dashed diagonal axis a per 
capita income increase greater than the national average. 



30 / Historical Review of Regional Economic Growth 




150 160 170 TW 190 200 



Figure 8. Relative State Economic Growth, 1939-54. 

Source: U. S. Bureau of the Census, Personal Income by States since 1929, 
Tables 1, 2, and 3. 



The groups of states comprising the first three categories have the common 
feature of an above-average increase in per capita income. In the three remain- 
ing categories, per capita income increased less rapidly than for the nation 
as a whole. 

IV. Above-average increases in population and total income; below-average 
increases in per capita income: The Far West falls into this category. Wash- 



Economic Growth, 1939-1954 / 31 



V19M 



i National average increase ^ } 
in per capita income S 

\ " | -| / 

j \ s\ \ \ \ 

-r~y J 

y 




& 



100 



no 



! I 
i 1 

! 123 




1 [ 


120 


1% 


140 150 


Inneose in 


population 


P1954 
Pli35 X1CO 



160 



170 180 W8 200 



Figure 9. Relative Regional Economic Growth, 1939-54. 

Source: U. S. Bureau of the Census, Personal Income by States since 1929, 
Tables 1, 2, and 3. 



ington was the only state in the region with a per capita income increase equal 
to the national average. 

V. Above-average population increases; below-average increases in total and 
per capita income: This category includes three small, largely urbanized 
states — New Jersey, Connecticut, and Delaware — and the District of Columbia. 
The District shows an actual drop in per capita income, evidently due to sub- 
urbanization and a change in the population structure of the city. 



32 / Historical Review of Regional Economic Growth 

VI. Below-average increases on all three counts: All the New England states 
fall into this category except Connecticut (which reflects the special situation 
of the suburbanization of New York City) . So also do the highly industrialized 
states, New York, Pennsylvania, and Illinois. On the basis of regional averages, 
the Middle Atlantic region is in this category; Maryland was the only state 
here that had an above-average increase in total income, a reflection in part at 
least of the suburbanization of Washington, D. C. 

Percentage-wise increases in per capita income, it can be seen, have not 
been at all closely correlated with growth in population. Thus, of the twenty- 
nine states in which per capita income increased more rapidly than the national 
average, nineteen had a below-average increase in population. Of the six 
states with the highest relative increase in per capita income — Kansas, Alabama, 
Oklahoma, Nebraska, Mississippi, and Arkansas, in that order — all but two 
showed an actual decline in population; and in these two — Kansas and Alabama 
— the population increase was below the national average. 3 

The discussion to this point has emphasized relative state and regional 
growth. However, absolute numbers must also be considered to obtain a 
rounded picture of regional population and economic growth. The "size" of 
the state or region is, of course, significant in this regard. Thus, for example, 
the states with the highest percentage increases in population between 1939 and 
1954, Nevada and Arizona, had total population gains of 105,000 and 446,000, 
respectively, while Pennsylvania, which ranked 36th among the states in 
percentage increase, had a population gain of 955,000. New York State, which 
was below the national average in percentage of population increase (117% 
against a national average of 123%), had the second largest absolute gain in 
the nation, a gain of 2,305,000. On the basis of absolute increments to popula- 
tion alone, the heavily populated states of Illinois, Massachusetts, New York, 
and Pennsylvania might be included among the "growth" states, even though 
each had a below-average percentage increase in population. A number of 
the more heavily populated states experienced sizeable population increases 
in both absolute and relative terms, particularly California, Florida, Texas, 
and Michigan. California's population gain was 5,723,000 out of a national 
total increase of 30,311,000 — almost one-fifth of the total. The increase for 

8 The relationship between population numbers and income is, of course, highly com- 
plex — and will be discussed in some detail in Part V of this book. At this point, however, 
it is appropriate to note at least one aspect of this relationship. Parts of the West, including 
California, Nevada, and Oregon, show a tendency for population increases to run stronger 
than increases in total income earned, so that increments in per capita income fail to keep 
pace with the national average. This does not necessarily mean a poor balance between 
population and income growth, however. As long as employment opportunities are at least 
at the national average, and wages remain at above-average levels for comparable work, it 
may be that the loss in relative per capita income standing of these states is due to the 
in-migration of lower-income earning families at the same time that the older residents — 
those that were there before 1939 — enjoy a rise in income at least comparable to the national 
average. However, the danger of imbalance does exist, and one would want to focus 
attention on changes in levels of real wages and similar indices to determine whether such 
imbalance has come into being. 



Economic Growth, 1939-1954 / 33 

Florida was 1,553,000; for Texas, 2,102,000; and for Michigan, 1,920,000. 

These two significant dimensions of population growth can be combined 
by employing the "shift" method of presenting data. 4 Because this method of 
measurement is used throughout the study, we shall explain it here at this point, 
where it is first employed. 



Population Shifts Among the States (Including a Note on Methodology) 

The net shifts by which we measure state and regional growth or decline are 
the relative gains and losses among the states with regard to a given variable 
(such as population growth, as here) in comparison to the national figures. 
These relative changes are determined in the following way. In measuring 
the population changes between 1939 and 1954, for example, the percentage 
increase in national population for these years is applied to each state's popula- 
tion to give what might be called the expected change by 1954 if population 
had increased uniformly throughout the states. This expected population 
change for each state is then compared with the actual population change in each 
state as estimated by the United States Census Bureau for 1954. The fact that 
population did not increase uniformly throughout the states means that some 
states achieved more and others less than the expected increase; i.e., some states 
increased their numbers proportionally more than the ratio over the 15-year 
period; others increased their number proportionally less than the ratio. For 
example, California's actual increase over the period 1939-1954 was 4,152,000 
more than it would have been if population had increased in the same propor- 
tion as the ratio (its expected change), whereas New York's increase in 



Population 
1939 
1954 

Actual Change: 

Expected Change: the percentage 
increase in national population 
applied to the states 

Net Shift Among States, Upward 
and Downward 

State's % of Total Upward or Down- 
ward Shift 



United States 
130,880,000 
161,191,000 


California 

6,785,000 

12,508,000 


"S o © 

^ co eo 

<h i— 1 t-i 


+30,311,000 


+ 5,723,000 


+2,305,000 




+ 1,571,000 


+3,132,000 


±9,458,000 


+ 4,152,000 


—827,000 




43.90% 


—8.74% 



4 Earlier studies which have employed the "shift" technique of organizing substantial data 
include Daniel Creamer in National Resources Planning Board, Industrial Location and 
Natural Resources (Washington: U.S. Government Printing Office, December 1943), 
Chapter 4; Wilbur Zelinsky, "A Method For Measuring Change in The Distribution of 
Manufacturing Activity: The United States, 1939-47," Economic Geography, Vol. 34 (April 
1958), pp. 95-126; Victor R. Fuchs, "Changes in the Location of U. S. Manufacturing 
Since 1929," Journal of Regional Science, Vol. 1 (Spring 1959), pp. 1-17. 



34 / Historical Review of Regional Economic Growth 




Absolute upward shift 
t: :-. :-:-.'| Relative downward shift 
| ^J Absolute downward shift 



Figure 10. Net Shift in Total Population, 1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 31.2. Total net shift as % of 1939 population: 7.2. 

Source: Appendix Table H. 



population was 827,000 less than it should have been to conform with the 
national increase. Thus, it can be said that California had a "net upward shift" 
of 4,152,000 people and New York a "net downward shift" of 827,000 over 
this period. California's share of the total net upward shift (that is, the sum of 
all state upward shifts) over the period was 43.9%, while New York sustained 
8.74% of the total net downward shift (Figure 10) . 5 

Figure 10 shows the net shift in population, by states, between 1939 and 
1954. The states shown in dark color are those which had net upward shifts; 
the others all had net downward shifts. The percentage figure shown for each 
state is the proportion which its actual number is of the total net upward or 
downward shifts. 

All states in the Far West and all except Oklahoma in the Southwest show 
relative population gains over these years. So also do the eastern Great Lakes 

5 Both the downward and upward shifts can further be classified as absolute or relative. 
An absolute downward shift is one in which all or some part of the downward shift was 
contributed by an actual decline in population. A relative downward shift is one in which 
there has been an increase in population, but one insufficient relative to the total increase 
to maintain that state's share of the total. An absolute upward shift is one in which all 
or some part of the shift was contributed by an actual increase in population. A relative 
upward shift is one in which there has been an actual decrease in the population but one 
that has been insufficient relative to the total decline to reduce that state's share of the 
total. 



Economic Growth, 1939-1954 / 35 

states and the Middle Atlantic seaboard states — Connecticut, New Jersey, 
Delaware, and Maryland; the gains in these Middle Atlantic states, as in 
Virginia in the Southeast, reflect the suburbanization effect noted earlier. 
Four states stand out: California, with 44% of the total net upward shift, 
Florida with 12%, Michigan with some 8%, and Texas with about 7%. 

The group of states along the Appalachians — Kentucky, West Virginia, 
Pennsylvania, New York, and Massachusetts — together account for 37% of 
the net downward shift; Illinois, Iowa, and Missouri for 16%; and a tier of four 
states extending from Alabama through Oklahoma for another 23%. Together, 
these three groups make up three-quarters of the net downward shift. 

In Table 10 the principal net population shifts among the states are examined 
for a somewhat longer time period — 1910-1957 — as a check on the findings 
for the shorter 1939-1954 period. The pattern of net upward shifts is generally 
similar for the two periods. The same four states — California, Florida, 
Michigan, and Texas — had the largest net upward population shifts between 



Table 10. Principal population shifts among the states, 1910-1957 





Percentage net 




Percentage net 




upward shift 




downward shift 


California 


43.11 


Pennsylvania 


14.28 


Florida 


12.24 


Missouri 


8.35 


Michigan 


11.77 


Massachusetts 


6.20 


Texas 


8.70 


Iowa 


5.98 


New Jersey 


4.20 


Kentucky 


5.44 


Arizona 


3.43 


Arkansas 


5.20 


Washington 


2.75 


Mississippi 


5.18 


Oregon 


2.37 


Georgia 


4.77 


Maryland-District 




Kansas 


4.51 


of Columbia 


2.25 


New York 


4.49 


North Carolina 


1.87 







Source: See Table 1, note. 



1910-1957 as between 1939-1954. These states are clearly the outstanding 
population-growth states of the twentieth century. 

With only slight modifications, also, the same groups of states had the greatest 
downward shifts in relative population standing in both periods. 6 

6 This play-back over time, while complicating the presentation, is important as a con- 
tinuing check on the growth patterns described, so that generalities with regard to growth 
experiences may be drawn with confidence. This is particularly important in the case of 
some series because, as mentioned at several points, beginning and/or terminal dates of an 
historical series can readily reflect the special circumstances of a single year or a very 
short period. 



36 / Historical Review of Regional Economic Growth 



Relative Total Income Gains and Losses 

When shift analysis is applied to the changes in relative standing among 
the states with regard to total personal income over the 1939-1954 period, 
California, Texas, Florida, and Michigan are found to have had the greatest 
net upward shifts. These are clearly the states that have made the greatest 
gains in volume of economic activities in recent years. On the other end of the 
scale, the four highly industrialized states of Illinois, Pennsylvania, New York 
and Massachusetts accounted for 80% of the net downward shift in total in- 
come. It should be emphasized again, however, that these were losses in rela- 
tive standing only. All four of these states were in the top quarter of the states 
ranked according to absolute increments in total personal income. Moreover, 
at the end of the 1939-1954 period, each of them had a larger proportion of 
the nation's total income than of its total population. 

The main divergences between changes in relative population standing and 
relative total income standing occurred in the Middle Atlantic seaboard states 
and in the Southeast (with the exception here of Florida and Virginia). Con- 
necticut, New Jersey, and Delaware all show a net upward shift in population 
but some loss of standing in total income. (As explained later, however, this 
arises from special circumstances surrounding the period itself and is not 
representative of the growth patterns of the area generally.) The Southeastern 
states improved their relative standing on total income between 1939 and 1954 
but had downward shifts in population. There was some divergence between 
population and total income changes in the Plains states also ; all of these states 
had net downward shifts in population, but a number of them (particularly 
Kansas and Nebraska) improved their income standing somewhat during this 
period. World War II had a profound effect on the agriculture of the South and 
the West and also on the industry and services of these areas to the extent that 
interior locations for war-connected activities were favored over seaboard loca- 
tions. An examination of the net shifts in total personal income for the postwar 
period alone shows that the main gains of these regions were made during the 
war and that the postwar pattern of income growth coincides fairly closely with 
the 1939-1954 pattern of population shift. For the postwar years, the Plains 
states show no gain in relative standing on total income; and this is true also 
for the states in the Deep South except Georgia and South Carolina. Further- 
more, for the postwar years considered alone, the Middle Atlantic seaboard 
states show a gain in income standing and in fact claim some 15% of the 
nation's net upward shift in total income. 

Table 11 shows the net upward and downward shifts in total income over a 
somewhat longer time period — the years from 1920 to 1957. Here the Middle 
Atlantic seaboard states — Connecticut, New Jersey, Maryland (and Virginia 
in the Southeast), together with the outstanding "growth states" of the shorter 
period examined — California, Florida, Michigan, and Texas — show up as hav- 
ing made the greatest improvements in relative position. Also, for this longer 
period, New York, Pennsylvania, Massachusetts, and Illinois show relative 



Economic Growth, 1939-1954 / 37 

losses as in the shorter period. These four states, together with four Plains 
states, had the greatest net downward shifts between 1920 and 1957. 

In general, then, it is clear that the World War II period has to be considered 
as special. The postwar period, however, exhibits the same pattern of differen- 
tial gains among the states as does the longer-run period of 1920-1957. 



Table 11. Principal shifts in total personal income among the states, 1920-1957 





Percentage net 




Percentage net 




upward shift 




downward shift 


California 


40.72 


New York 


26.51 


Florida 


12.15 


Pennsylvania 


19.76 


Texas 


8.77 


Massachusetts 


13.32 


Michigan 


8.16 


Illinois 


6.85 


Virginia 


3.35 


Iowa 


3.72 


North Carolina 


3.20 


Missouri 


3.55 


New Jersey 


3.11 


Oklahoma 


3.09 


Maryland-District 


* 


Kansas 


2.99 


of Columbia 


2.30 


Nebraska 


2.10 


Louisiana 


2.25 


Rhode Island 


1.85 


Connecticut 


2.13 







Source: See Table 8, note. 



Per Capita Income 



The evolving balance between population numbers and total income received 
is highly suggestive of the type of economic adjustment being made within a 
given state or region. Comparative percentage gains among the states in per 
capita income have already been presented. Figure 11 provides a useful addi- 
tional view of the changing income picture. It shows the absolute increments 
to real per capita income from 1939 to 1954. This additional view is needed 
in order to avoid overstating the welfare advances of the areas starting with 
relatively low per capita incomes and understating the real advances of the 
"richer" and particularly the middle-income states. 

Thus, the Southeast as a whole has had percentage increases in per capita 
income greater than the national average for every decade except one since 
1900; yet, during the same period, there has been only one decade in which 
absolute increments in per capita income have exceeded the national average in 
any Southeastern state except Florida. This important feature of economic 
growth can be highlighted through an example, taking the richest of the 
Southeastern states. Between 1880 and 1950, the percentage increase in real 
per capita income for Florida was well above the national average; for every 
decade but one, the percentage increase for New York has been less than the 



38 / Historical Review of Regional Economic Growth 

national average. Yet in 1950 the absolute amount by which New York's per 
capita income exceeded that of Florida was $100 greater than it was in 1880. 
Of course, low income regions that realize a percentage rate of growth in real 
per capita income in excess of the national average over a sufficiently long 
period of time may ultimately liquidate the discrepancy in income. The start- 
ing point and rate of gain must, however, be considered to appreciate the length 
of time which might be involved in such equalization. 

The regional pattern shown in Figure 11 differs from the one that emerges 




IH Above U.S. average 
T^l Below U.S. average 



Figure 11. Absolute Increments to Real Per Capita Income, 1939-54. 

Source: U. S. Bureau of the Census, Personal Income by States since 1929, 
Table 2. 



when either levels of per capita income or percentage rates of increase in per 
capita income are recorded. And it is significantly different in certain respects 
from the pattern which emerges when growth in population and total income 
are set down. Possibly the most interesting feature of this particular measure 
of improvement in material welfare among the states, however, is the fact that, 
in absolute terms, the gains over the period 1939-1954 have been within a rela- 
tively narrow range. In only five states were the absolute gains in real income 
as much as 25% below the national average; and there were only two states 
with gains as much as 25% above the national average. This suggests a re- 
markably uniform forward movement in absolute per capita income gains across 
the entire country during this 15-year period. I The meaning of this forward 
movement is discussed in some detail in Part V of this book.) 



Economic Growth, 1939-1954 / 39 



Growth Profile Combining Multiple "Dimensions," 1939-1954 

In presenting measures of regional growth, we have touched upon six 
"dimensions" — absolute increases in population, in total real personal income, 
and in per capita real income; relative increases in population, in total real 
personal income, and in per capita real income. None of these measures alone 
is adequate to describe even the crude outlines of economic growth. Each is 
subject to certain limitations and biases. Absolute measures tend to overstate 
the growth of big states such as Pennsylvania and understate the growth of 
little states such as Delaware. Relative measures tend to overstate the growth 
of areas with a low population or small income base and understate the growth 
of areas with a high population or large income base. Total income and popu- 
lation increments cannot indicate changes in individual and family welfare; 
and per capita income increments frequently do not correlate with other dimen- 
sions of growth. 

In a situation such as this, one is tempted to try to combine the different 
variables into a combined index of growth. There is no possibility, however, 
of ever knowing the proper weights for combining these elements or the proper 
way to bring about an offsetting of bias. On the other hand, there does seem 
to be some advantage in bringing these measures together state by state into a 
kind of profile that will allow a comparison of the similarities and diversities 
that exist. 

To this end, Figure 12 has been prepared. In it, we resort to the use of 
symbolic classification. Each state is assigned a plus or a minus for each of the 
six measures according to whether it ranks above or below the national average 




\,/\ Stronger in absolute measures 
. I Stronger in relative measures 
I Relatively little growth, if any, during 1939-54 




States with an equal division of plus sians 
Wfi are classified according to 'he general 

tendency for the reaion. 




Figure 12. Profile of Growth Dimensions, by State, 1939-54. 



40 / Historical Review of Regional Economic Growth 

or median for that measure. The symbols on the map read from left to right 
for (1) population, (2) total income, and (3) per capita income. The top row 
represents the relative measures ; the bottom row, the absolute measures. 7 

Even as represented by these limited and relatively crude indicators, the 
states exhibit a wide variety of growth patterns. At one extreme are Texas, 
Michigan, Indiana, and Washington, that are above the average for every 
dimension; at the other extreme are Maine, New Hampshire, Vermont, Rhode 
Island, and West Virginia, that are below the average for every dimension. 
The mature industrial states of the Middle Atlantic region and the western 
Great Lakes states tend to be strong in absolute dimensions and weak in rela- 
tive dimensions. In contrast, the agrarian states in the Southeast and Plains 
regions and the agrarian-mining states in the Southwest and the southern 
Mountain region tend to be strong in relative dimensions and somewhat 
weaker in absolute dimensions. The northern tier states in the New England, 
Plains, and Mountain regions tend to be weak in both. 



A Sub-State View 

Our investigation in this study is based mainly on state and regional data. 
However, at a number of points, we have examined data on sub-state areas, 
particularly the metropolitan communities and agricultural groups of counties. 
We have done so, first, as a check on whether the intra-state variations are too 
large to permit a meaningful description and analysis of the development of 
the sub-national economy and, second, as a way of enriching the information 
available, particularly in discussing economic activities that tend to be highly 
localized. 

The smaller area figures that we have used are Bureau of Census data on 
the so-called State Economic Areas. These areas are groups of counties — 
either metropolitan or non-metropolitan — with similar economic and social 
characteristics. 8 

The extent to which these data can be used in checking the state view of 
income levels and economic growth is limited by the fact that they differ from 
the data available for the states both as to the measures employed (and the 
nature of the basic data) and as to the years covered. However, some mean- 
ingful comparisons are possible. Because of our concern with income levels, 
it is particularly important to compare the state and sub-state income figures 
to see whether the state and regional view of this critical variable is actually 

'The method used has the obvious weakness of disregarding how much a state is above or 
below the average or median, but it did not seem advisable to elaborate the method. The 
only purpose is to present a simple impressionistic generalized picture of relative growth. 

8 These data have been brought together by Otis Dudley Duncan and his associates. The 
SEA system for which data are given here is based on State Economic Areas by Donald J. 
Bogue (Washington, D. C: U. S. Government Printing Office, 1951). While some of these 
data are presented in the present volume, most of the sub-state material is presented and 
discussed in the volume Metropolis and Region (Baltimore: The Johns Hopkins Press, 
1960) , which is under the authorship of Duncan and his associates. 



Economic Growth, 1939-1954 / 41 

too gross for purposes of analysis. Such a comparison can be made for one 
year, 1949, when census data on median family incomes are available for State 
Economic Areas. 9 The data for that year are presented in Figure 13. This 
chart effectively serves the two purposes referred to above. First, it suggests 
that the state and regional figures on income levels — as shown in Figure 7, for 
example — do as a matter of fact provide a good general description of income 
levels in the sub-national economy. There is clearly a strong tendency for levels 
of living to be similar over wide stretches of territory. The Southeast, in this 
view as in the state view, is clearly in the lowest income category. The Manu- 
facturing Belt and the Far West are seen to fall in the highest income classes, 
while the other three western regions generally fall into a middle income cate- 
gory. 10 In other words, it becomes evident that the state figures do describe 
the over-all income situation quite well. 11 

The sub-state data are particularly valuable from the standpoint of permitting 
a separate view of the urban and agricultural areas within the various states. 
How do levels of living and growth patterns differ as between these two major 
types of areas? The data in Figure 13 are revealing on this score. We can see 
that, in the richer parts of the nation — particularly in the Manufacturing Belt 
and the Far West — the differences in income levels as between city and country- 
side are not nearly so great as are the differences among the multi-state regions. 
In these richer sections, both non-metropolitan and metropolitan State Economic 
Areas fall within the highest income category. 12 In the middle-income regions, 
the urban areas tend to fall into the highest income category (highest, that is, 
when all the SEA's are classified in quintiles), while the rural zones tend to 
fall into somewhat lower categories. In the lowest income region, the Southeast, 
the urban areas tend to enjoy higher levels of living than the farming areas, 

9 Data were obtained from an unpublished series of data for State Economic Areas pre- 
pared by Donald J. Bogue and Calvin L. Beale and made available for the present study. 
Median family income for the SEA's has been determined on the basis of income distribu- 
tions for the SEA as a whole. It has not been estimated by averaging median family incomes 
of the counties making up each SEA. In those cases where there were differences between 
the composition of SEA's used by Bogue and those used by the present authors, adjustments 
were made by recomputing median family income using the 1950 U. S. Census of Population. 

10 The high incomes shown for the Mountain SEA's for 1949 are not representative of 
income levels for this part of the nation generally. They reflect the unusually high farm 
prices and incomes which characterized the immediate postwar years. In a longer run view, 
the Mountain areas clearly fall into the middle income category. 

11 At the same time, the sub-state data provide useful additional information which is 
suggestive for a fuller view of sub-national economic growth. Thus, it can be seen from 
Figure 13 that Northern New England, the Northern Lakes area, and the Appalachian area 
of Pennsylvania and New York tend to have lower income levels than does the remainder of 
the Northeast. On the other hand, the lower Appalachian area, in the Southeast — which, as 
will be shown later, contains the greatest concentration of manufacturing activity in the 
Southeast — has a higher income level than the remainder of that region. The relatively wide 
range of income levels among the SEA's of Florida, Louisiana and Texas is suggestive of 
the significant economic transition that these "in-between" states are experiencing. Once 
brought to the forefront, sub-state materials of this kind can readily be taken account of in 
analyzing the differential growth of the various parts of the nation. 

"Excepting only the cities in the "depressed" areas. 



42 / Historical Review of Regional Economic Growth 




Economic Growth, 1939-1954 / 43 

but average income levels of the Southern cities are, on the whole, below those 
of cities in other parts of the country. 

The sub-state data indicate that the growth of urban population, as well as 
average levels of living, has followed discernible regional patterns (but always 
with significant within-region variations). Figure 14 brings together data on 
the percentage increases in population for the larger metropolitan communities 
(actually Standard Metropolitan Areas, which are essentially the same as the 
metropolitan State Economic Areas) for the decade 1940-1950 and the 1949 
data on median family income within metropolitan communities. When the 
data for these two variables are plotted on the same chart, there is a definite 
clustering which permits the drawing of boundary lines representing the 
"greater regions" — that is, the Northeast, Southeast, and West. 

While the patterns are not entirely neat (the SMA's of the Plains region in 
particular tend to be scattered), the picture that emerges is quite similar to 
the one presented by the state data. The metropolitan communities of the 
Northeast have had far less population growth than those in the West and the 
Southeast. Within the Northeast, as marked off in the chart, the cities that 
have had an above-average growth are almost entirely within the Great Lakes 
region. For the Northeast, the range of percentage increases in population for 
the decade 1940-1950 has been relatively narrow, while income levels extend 
over a rather wide range. By contrast, the Western metropolitan communities 
fall within a narrow income range, while their population growth rates extend 
from a little above the median for the nation to truly phenomenal rates of 
growth (the population of Albuquerque, New Mexico, more than doubled 
within the decade). Income in the Southeast metropolitan communities is 
below the median for the nation with only one exception (Richmond, Virginia). 
While some southeastern cities have had a sizable growth in population, the 
greatest number outside of Florida experienced a rate of growth of modest 
dimensions, though largely above the average for the nation. 

For the farming sections, sub-state data on levels of living are available over 
a fairly long period of time. A Farm-Operator Family Level-of -Living (FOLL) 
Index covering the State Economic Areas is available for 1930, 1940, 1950 and 
1954. 13 Farm levels of living at the two ends of this period as shown by this 
index are set down in Figure 15 (for 1930) and Figure 16 (for 1954) for the 

"This index, developed by Margaret Jarman Hagood, is a composite of several differently 
weighted variables related to levels of living: (1) percentage of farms with electricity; 
(2) percentage of farms with telephones; (3) percentage of farms with automobiles; and 
(4) average value of products sold or traded in the year preceding the census (adjusted for 
changes in purchasing power of the farmer's dollar). See Margaret Jarman Hagood, Farm- 
Operator Family Level-of -Living Indexes for Counties of the United States, 1930, 1940, 1945, 
and 1950 (Washington: Department of Agriculture, Bureau of Agricultural Economics, 
1952). 



Figure 13 — opposite. Median Family Income, by State Economic Areas, 1949. 

Source: An unpublished series of data for state economic areas prepared 
by Donald J. Bogue and Calvin L. Beale. 



44 / Historical Review of Regional Economic Growth 




.. • Altoona, Pa. 

i- ,,::©uluth, Minn.-Superior, Wise. 

I w Johnstown, Pa. • ,.«;;* ' 

*•" Wheeling, W.Va.| 

Steubenville, Ohio| 

I 
Wilkes-Barre-Hozleron, Po. » 

!i 

I 

I I 

200C 2250 2500 2750 3305 3250 3500 3750 



• NortheoH N»* England Middle Arionhc and Greoi lak*v 
m Werf (Far West Mountain and Southwest*} 



: 



*£xc/udf« 12 SAW's m Pteins /?egt< 



11II1 



Figure 14. Regional Patterns of Family Income (1949) and Population 
Growth (1940-50) for SMA's of 100,000 or More Population. 

Source: U. S. Bureau of the Census, 1950 Census of Population, Vol. 11, 
Pt. 1, Tables 86 and 92. 



Economic Growth, 1939-1954 / 45 

non-metropolitan SEA's. During this period, the levels of living of farm opera- 
tors had undergone a significant increase in every part of the country. And 
yet, as is clear from these charts, the relative position of areas with respect to 
levels of living of farm populations has remained much the same for nearly a 
quarter of a century, if not longer. Regional differentials in 1930 and 1954 
are strikingly similar. What has happened over the period is spelled out in 
Table 12, which shows the means and standard deviations of the FOLL indexes 
for the non-metropolitan SEA's. 

Table 12. Means and standard deviations of farm-operator family level-of- 
living indexes for 348 non-metropolitan state economic areas, 1930, 1940, 1945, 
1950, and 1954 

Year Mean Standard deviation 

1930 76 36 

1940 81 38 

1945 102 44 

1950 123 40 

1954 142 36 

Source: M. J. Hagood, Farm-Operator Family Level-of-Living Indexes for Counties of the 
United States, 1930, 1940, 1945, and 1950 (Washington: Department of Agriculture, 
Bureau of Agricultural Economics, 1952) ; and M. J. Hagood, C. K. Bowles, and R. R. 
Mount, Farm-Operator Family Level-of-Living Indexes for Counties of the United 
States, 1945, 1950 and 1954 (Washington: Department of Agriculture, Agricultural 
Marketing Service, 1957). 



The dispersion of FOLL values around the mean, as measured by the standard 
deviation, has remained remarkably constant over the interval. The equaliza- 
tion in levels of living among regions that has taken place has clearly come 
about because of an over-all increase in levels of living throughout the nation 
and not because of much more rapid increases specifically taking place in the 
poorer areas. Greater increases in levels of living in the poorer areas would 
have brought about an increase in the mean FOLL index, but the standard 
deviation would have been reduced. This has not been the case. Relative posi- 
tions of the various rural sections of the country have remained essentially 
the same. 

In spite of some significant intra-state variations, the picture that emerges 
is quite similar to the one that was highlighted by the state and regional statis- 
tics. This brief excursion into the sub-state view, while it provides some en- 
riching information, strongly suggests that sub-national economic growth can 
be described (and analyzed) meaningfully by focusing mainly on state and 
regional data. In the course of our study, our examination of the sub-state data 
which was much more extensive and detailed than the data presented here 
would reveal, gave us the assurance that it was sensible to center attention 
mainly on developments as revealed by the state data. 



46 / Historical Review of Regional Economic Growth 




Economic Growth, 1939-1954 / 47 



Summary of Regional Economic Development, 1870-1957 

The examination of growth indexes covering multi-state regions, states, and 
sub-state units for almost a century, as well as for short recent periods, provides 
a highly suggestive overview of economic growth in the United States. This 
overview can be summarized as follows: 

Regional economic growth over the past century can be viewed appropriately 
only against the background of the phenomenal growth of the national economy. 
The population of the nation grew from 40 millions in 1870 to 170 millions in 
1957. Total income of the nation increased in current dollars from less than $9 
billion in 1880 to some $345 billion in 1957; even if account is taken of price 
increases, the growth in income has been truly remarkable. The gain in material 
welfare which has taken place because of the far more rapid growth in income 
than in population is suggested by the increase in real income per capita (in 
1929 dollars) from $302 in 1880 to $1,236 in 1957. Even though at best this 
is a very rough estimate, particularly in consideration of the great changes in 
the "basket of goods" and of services consumed, the general order of magnitude 
is not likely to be too far off the mark. 

Since 1870, the population of the United States has spread westward across 
the continent to occupy the remaining agricultural land and to tap other 
opportunities of the West. After 1910, however, the westward movement 
slackened until after World War II, when it was again revised mainly by the 
striking growth of California. In the period between 1870 and 1910, the 
proportion of the nation's population in the four Western regions increased 
from some 14% to 26% of the total; by 1957, it had gone up to 30% — that is, 
only 4 percentage points above the figure for 1910. 

In absolute numbers, the steady growth of population east of the Mississippi 
has been unbroken in spite of a continuing exodus from northern New England 
and some parts of the South. After 1920, in fact, the greatest relative and 
absolute losses in population were recorded in the Mountain and Plains regions 
as people moved westward into California and some parts of the Southwest or 
eastward into the industrial areas of the Great Lakes and Middle Atlantic 
regions. In recent decades, a movement of population into Florida has become 
marked. 

Since 1910 almost all interstate and intrastate migration has been into urban 
areas. Urbanization has continued, although at a declining rate, and the 1950 
census reported that 64% of the entire national population resides in cities. 14 
Only less significant than urbanization itself has been the growth of large and 
medium-sized cities which have come to dominate the pattern of settlement and 

"Some 60% according to the older definition of urban. 



Figure 15 — opposite. Farm-Operator Family Level-of-Living Indexes for 
Nonmetropolitan Areas, 1930. 



48 / Historical Review of Regional Economic Growth 




Economic Growth, 1939-1954 / 49 

activities in their respective regions. Only 18 states in 1950 had more than half 
their population still living in rural areas; of these one was in New England, 
three were in the Mountain region, four in the Plains, and ten in the Southeast. 

The growth in total personal income — a key measure of the expansion in 
volume of economic activities — has followed a pattern essentially similar to that 
of population growth as far as the East-West division is concerned. As recently 
as 1880, more than 80% of the total national income was received by the four 
Eastern regions (which contained at that time 81% of the total population). By 
1920, the proportion had fallen to a little over 74%, most of the change taking 
place before the turn of the century. By 1957, some 70% of total personal 
income was still received by the Eastern regions; this was almost exactly the 
same proportion of the national total as in the case of population. 

The record of relative regional growth lends itself to a choice of emphasis — 
either upon the West's remarkable growth or upon the East's equally remarkable 




Percent of total mfg. employment 74% j 64%* 



Percent of total personal income 



* The figure is for J 956 

Figure 17. The Manufacturing Belt, 1900 and 1957. 



maintenance of national dominance, especially that of the Manufacturing Belt. 
As Figure 17 shows, after more than a half-century of "filling in" and probing 
of new wealth and new opportunities across the vast continent, the Manufac- 
turing Belt (covering roughly the three Northeastern regions excluding northern 
New England) in 1957 contained 46% of the total population of the nation, 



Figure 16. Farm-Operator Family Level-of-Living Indexes for Non- 
metropolitan Areas, 1954. 



50 / Historical Review of Regional Economic Growth 

53% of the total income, and 64% of the total manufacturing employment. 15 
Something over a half century ago, in 1900, the figures for the Manufacturing 
Belt were 49% of the nation's total population, 61% of the total income, and 
74% of the manufacturing employment. In one sense, then, it can be said that 
there has been a remarkably small loss in relative standing over a half century of 
great national growth, and clearly this area is still the very heart of the national 
economy and the very center of the national market for goods and services. 

On the other side, the gains of the remainder of the country, particularly of 
the Far West and the Southwest, are also strikingly demonstrated by these 
figures, and one can understand why the growth of the West should be made the 
central feature of many of the accounts of relative regional growth. 

It seems sensible, we believe, to give full attention to both of these views of the 
same phenomenon and to weigh them properly in the larger context of over-all 
national economic growth. 

When the focus is narrowed down from this very broad view to one encom- 
passing the individual states (and groupings of states), and particularly when 
attention is devoted to the more recent decades, the data reveal several types of 
situations with regard to growth in volume. These can be briefly characterized 
as follows: 

I. Four regions and one state can be identified as having had the greatest 
relative over-all expansion in population and economic activities generally in 
recent years. These are: the Far West, the Southwest, the eastern Lakes states, 
the Atlantic Seaboard states, and Florida. 

II. The picture for regions of greatest relative decline — or least differential 
growth — is less clear-cut. There appear to be five groups of states in this cate- 
gory, some of which must be defined in special regional terms: 

(1) New England (other than Connecticut) ; 

(2) New York, Pennsylvania, West Virginia, and to a lesser extent Kentucky, 
which, together with Massachusetts, make up a band of states along the Appa- 
lachians which have declined in relative economic standing in the nation; 

(3) The "Deep South" states of Mississippi and Arkansas, which, together 
with Oklahoma, had absolute losses in population over the period 1939-1954 and 
only slight gains in volume of economic activities relative to the rest of the 
nation; 

(4) The northern Mountain states; and 

(5) The Plains states. 

These groups do not show decline on all scores by any means. The older and 
economically more advanced areas, in particular, rank high in absolute increases 
in population and total income. Excluding those states that show greatest abso- 
lute growth, four New England states, four northern tier states in the Mountain- 
Plains section of the country, the "Deep South" states, and West Virginia are 
the areas that show the least economic growth over the recent past. (Refer back 
to Figure 12.) 

15 The figure for manufacturing employment is for 1956, the latest available at the time 
of writing. 



Economic Growth, 1939-1954 / 51 

On the side of material welfare, as measured by per capita income levels and 
rates of growth, several features stand out from the preceding review: 

1. All parts of the country have gained from the great increases in income 
that have accompanied the economic development of the United States; but in 
spite of some equalization, the three regions that were below the national 
average in per capita income in 1880 (the Southeast, Southwest, and Plains) 
were still below the national average in 1957 — the poorest region, the South- 
east, by 30%. 

2. There has been a clear trend toward income equalization throughout the 
entire period. The most striking feature of this trend, however, has been a 
strong "pulling down" towards the national average of the highest income levels 
(mainly by a great inflow of population) . The "upward push" of the low income 
regions has been much less strong. Thus, while real per capita income of the 
Far West was 211% of the national average in 1880, it was 135% of the average 
in 1920 and only 119% in 1957 — even though it was still the highest per capita 
income of any region in the country. In the Mountain region, per capita income 
was 166% of the national average in 1880 (the second highest in the nation), 
102% in 1920 (slightly above the national average) but only 92% in 1957 
(below the average). On the other side, per capita income in the Plains was in 
the same relationship to the national average in 1957 (90%) as in 1880; the 
Southwest improved its relative position from 61% of the national average in 
1880 to 86% in 1957 and the Southeast moved up from 50% in 1880 to 70% 
in 1957. 

3. Relative levels of per capita income — again, in spite of the tendency 
toward equalization — have become strongly associated with the level of urbani- 
zation-industrialization, much more so than in the past. The highly industrial- 
ized and urbanized Middle Atlantic and New England regions have been well 
above the national average level of per capita income throughout the period 
since 1880, and the Great Lakes region joined their ranks as its industrialization 
gathered momentum. However, in the 1880-1900 decades, the two regions with 
the highest per capita incomes — the Far West and Mountain regions — were not 
industrialized. In the 20th century, the development of these two regions 
diverged. 

After 1900 the Far West grew more in accordance with the pattern of the 
older urban-industrial regions of the East. The Mountain region, in contrast, 
declined steadily in relative levels of living, falling below the national average 
after 1920 and closely paralleling the agricultural regions. At some point 
between 1900 and the outbreak of World War I, the Far West appears to have 
crossed the threshold of urban-industrial society. In recent decades there has 
been a consistent division between the highly industrialized and the less indus- 
trialized regions with regard to relative levels of living. 

4. The data, particularly those for the recent years, highlight the fact that 
there have been a variety of paths to increases in levels of per capita income. 
Some states and regions have experienced above-average increases in income 
either in relative or absolute terms, or both, while they have grown in population 
and economic activities. This has been the case with most of the Far West and 



52 / Historical Review of Regional Economic Growth 

the Southwest, Florida and Virginia, and the eastern Great Lakes states. On the 
other side, some of the states have had gains in material welfare in the face of 
out-migration and relatively little over-all increases in population and economic 
activities, as is true of the Plains region generally and some of the Southeastern 
states. In a few instances — particularly Oklahoma, Arkansas, and Mississippi — 
there has been a substantial relative gain in per capita income levels accom- 
panying an actual decline in population. Clearly, "sound economic adjustment" 
has a different meaning under different circumstances. 



Part 
J J A Framework for Analysis 



^ Theories of regional economic growth 



The growth of economic activity and of per capita income, we have seen, has 
differed widely among the various states and regions over the years. We shall 
now attempt to provide a conceptual framework within which the main elements 
in this growth phenomenon can be described and analyzed. 

To begin, let us consider for a moment how our economic system determines 
what commodities are produced in which locations and in what amounts and 
who receives income and in what amounts. In our society it is the market mech- 
anism or price system that organizes and directs economic activity. The prices 
of final products and of production factors in different places, as well as the 
quantities of final products produced and of production factors employed in 
these places, are arrived at through a process of mutal determination often 
referred to as the general equilibrium system. Through its workings, the volume 
and composition of output and the factor prices in different places are deter- 
mined. The factor prices — essentially, the prices of labor and capital — together 
with the pattern of ownership of these factors determine the income of different 
persons. 

The basic data for economic analysis of the workings of this system are data 
on number, tastes, and location of consumers; quantity and location of all 
factors of production; technological developments; and institutional influences. 
As the basic elements change over time, the system generates new solutions to 
the economic problem which involve changes in the volume and composition of 
output in different places and in the incomes of persons in these places. 

It is not our intention here to attempt a study of the workings of the general 
equilibrium system in its entirety. Rather, as in any applied research project in 
economics, we shall seek to identify those variables within its operation that are 
of strategic importance for the particular problem with which we are concerned 
— the problem of determining relative growth in the volume of economic activity 
and in per capita incomes or welfare in different places. 



Volume and Welfare Aspects of Growth 

Growth in total output, we have shown, is not always associated with growth 
in per capita income. For this reason we shall consider these two kinds of 



/ 55 



56/^4 Framework for Analysis 

growth phenomena separately in our analysis. Both the volume and welfare 
aspects of economic development, it is true, depend generally on the same set of 
forces; but the strategic variables that determine them need not be identical. 
The forces determining the former — the "volume" forces — are those that affect 
relative costs of production (both costs of material inputs and labor costs) in 
the various parts of the country, the relative size of different market centers 
(both final or consumer and intermediate producer markets), and relative trans- 
port costs as between different places. The forces determining the level of per 
capita income, on the other hand — the "welfare" forces — are those that influence 
the relation between the rate of increase in population and labor force and the 
rate of increase in non-human factors of production within given regions and 
those that differentiate one region from another in the characteristics of the 
population — in their skills, age, sex, and racial composition. 

There is another reason also why the volume and welfare aspects of regional 
economic growth should be considered separately. Economic efficiency is widely 
accepted in the United States as a desirable social goal. Technically, economic 
efficiency means an organization of production and distribution of the national 
product so that the highest total real value of output is achieved with the 
resources available. The pertinent point here is that national economic efficiency 
may require that the volume of economic activity in certain areas grow less 
rapidly than in others or actually decline. 

To illustrate, let us consider the hypothetical instance of an area which pro- 
duces nothing but a single agricultural commodity under competitive conditions, 
and which is suitable for no other kind of production. Let us suppose further 
that, as economic development proceeds, the demand for this commodity 
declines. If production remains unchanged, the price of this commodity will 
then fall relative to the prices of other commodities, and the returns to factors 
of production employed in producing it will fall relative to the returns to identi- 
cal factors employed in producing other commodities elsewhere. Clearly, this is 
inefficient from the economic standpoint, since a transfer of labor and capital 
resources to other areas would increase the national product. If enough labor 
and capital factors migrate from our hypothetical area to other areas in response 
to their now lower incomes, 1 the incomes of these factors can be restored to 
equality with those in other areas and economically efficient production restored. 
(There may, of course, be important non-economic considerations which would 
make it desirable to subsidize a situation that is less than optimal from the 
economic standpoint alone. But under such circumstances it is important to 
understand just what is involved and what the economic costs actually are.) 

In general, then, the failure of the "volume" of economic activity to keep pace 
with the national average growth in certain areas may be helpful in achieving 
the socially desirable goal of economic efficiency. Yet this need not imply, as 
will be shown later, that the incomes of persons in the area affected — either 

2 In practice this would be accomplished by a migration of persons and a failure to replace 
the non-human agents of production as they wear out. To the extent that there are factors 
such as land which are physically fixed and do not wear out, such inputs would, of course, 
earn permanently lower incomes. 



Theories of Regional Economic Growth / 57 

those who migrate elsewhere or those who remain— need increase any less 
rapidly than the incomes of persons elsewhere, at least in the long run. 

Here we have a truly critical feature of our economic system: every part of 
the country may not have an equally rapid growth of economic activities, but 
persons in every part of the country, under the proper circumstances (e.g., 
prejudice aside), can enjoy like income for like work. 



Economic Growth Theories 

Two major theories have been advanced in recent years to explain subna- 
tional growth phenomena — or, possibly more accurately put, two major concepts 
of (or approaches to) the organization of aggregate data to help explain these 
phenomena have been proposed. These are the "export-base" concept and the 
"sector" concept. Both provide highly useful insights, but they are both 
basically limited in scope. 

THE EXPORT-BASE THEORY 

This theory, whether used to explain growth of regions or of cities, hypo- 
thesizes that the factor initiating growth and determining its extent is the 
"export-base." 2 Growth in a given unit, it is proposed, is initiated by the 
response of the industries within this unit to an increase in demand arising 
outside the unit itself. This results in an expansion of economic activities, 
particularly local trade and service activities, through a multiplier process 
similar to the familiar investment-multiplier and foreign-trade multiplier in 
national income models. 3 

Harold Innis and G. M. Meier, 4 in their studies of the growth of Canada's 

2 See Douglass C. North, "Location Theory and Regional Economic Growth," Journal of 
Political Economy, Vol. 63 (June 1955) pp. 243-58. The main development of the theory 
has been in terms of urban growth. A rich literature has grown up around what has come 
to be known as the theory of the Urban Economic Base. See the series of articles by 
Richard Andrews appearing in Land Economics beginning with the issue of May 1953; also 
articles in this journal and in the Journal of the American Institute of Planners by Charles 
M. Tiebout, John Alexander, Charles Leven, Hans Blumenfeld, Ralph W. Pfouts and others. 

3 Central to the export-base theory is the distinction between export industries and activ- 
ities and "residentiary" industries and activities. In the development of the theory with 
specific reference to urban communities, economic activities directly related to the export 
of goods and services to other areas or in other ways involving the earning of income from 
outside the area (such as profits from outside investment) have come to be termed "basic" 
activities and are said to make up the "economic base" of the metropolitan community. 
Other activities, such as local services and retail trade, are termed "nonbasic" or dependent 
activities, since their level is set by the income earned through the basic activities. 

"Harold A. Innis, Problems of Staple Production in Canada (Toronto: The Ryerson Press, 
1933) ; The Cod Fisheries: The History of an International Economy (New Haven: Yale 
University Press, 1940) ; G. M. Meier, "Economic Development and the Transfer Mech- 
anism: Canada, 1895-1913," The Canadian Journal of Economics and Political Science, Vol. 
19 (February 1953) pp. 1-19. 



58 / A Framework for Analysis 

national and regional economy, and Douglass C. North, 5 writing about regional 
economic growth in the United States, have stressed the key role of exportable 
commodities and services. Their position is that capital investment will tend 
to flow into a region to develop the export industries, including the improve- 
ment of production processes and further development of specialized services 
to the exports. The resulting increase in income will tend to augment demand 
for secondary products and to induce investment in a variety of other indus- 
tries. North states : 

"The social overhead benefits that have been created . . . and the develop- 
ment of a trained labor force and indigenous capital make it far easier to 
develop new exports. Whether such industries were originally residentiary 
and, by gradually overcoming transfer-cost disadvantages, become export in- 
dustries, or were originally footloose industries not significantly affected by 
transfer costs, the result is to broaden the export base. As such a region 
matures, the staple base will become less distinguishable, since its production 
will be so varied." 6 

The rate at which a region grows, it is suggested, will depend on the rate at 
which the export base expands in line with the increase in the demand for the 
region's exportable commodities and services. 

The export-base theory is particularly valuable in bringing to the forefront 
the fact that the growth of any subnational unit — whether a metropolitan com- 
munity or a large multi-state region — is directly tied to developments within 
the national economy and, in some cases, to changes in international trade 
as well. Because attention is focused mainly on the relationship between a single 
subnational unit and the "outside world" treated as a whole, the functional ties 
among regions are hidden from view; however, the changing patterns of na- 
tional demand and investment receive an appropriately central position in the 
analysis of regional growth. Aggregation of the industries and activities in a 
region in terms of export versus local use clearly provides a classification which 
is meaningful for understanding and measuring certain aspects of growth. 

THE SECTOR THEORY 

The economic-sector theory flows from the empirical observations of Colin 
Clark and Allan G. B. Fisher that a rise in per capita incomes in different 
countries or at different times is generally accompanied by a decline in the 
proportion of the labor force employed in agriculture and a rise, first in the 
proportion of employment in secondary activities (principally manufacturing), 
and then in tertiary or service activities. 7 The main reasons for this shift in 
the relative importance of the different sectors are different income elasticities 

5 Op. cit. 

Hbid., p. 256. 

7 Colin Clark, The Conditions of Economic Progress (London: Macmillan, 1940) ; Allan 
G. B. Fisher, "Capital and the Growth of Knowledge," Economic Journal, Vol. 43 (Sep- 
tember 1933), pp. 379-89 and "Production, Primary, Secondary and Tertiary," Economic 
Record, Vol. 15 (June 1939), pp. 24-38. 



Theories of Regional Economic Growth / 59 

of demand for their products and differential rates of change in labor produc- 
tivity. For example, demand for food and other agricultural products will 
normally increase less rapidly than per capita income, and at the same time 
farm output can be increased with a decreasing proportion of the labor force 
because of productivity gains through the use of machines, fertilizers, and im- 
proved techniques. 

Thus, the sector theory focuses on internal rather than external development; 
economic growth is seen as primarily an internal evolution of specialization 
and division of labor, although external shifts in demand are not ruled out 
as of no importance. 

As applied to regional economic growth, sector shifts are seen by the pro- 
ponents of the theory as providing the main dynamic of economic advance 
both in terms of growth in the volume of economic activities and in terms of 
improvements in per capita income (although normally the distinction is not 
clearly made). 8 

The sector theory, joined with suggestions from the theory of the location 
of economic activities, has_given rise to what might be termed a theory of 
development stages, that is, a theory regarding the "normal" sequence of 
stages through which regions experience economic growth. 9 It is suggested 
that the development of most regions will be characterized by the following 
sequence: (1) a stage of self-sufficient subsistence economy; (2) growth 
through production specialization in primary activities and interregional 
trade — which can be expected to accompany improvements in transport; (3) 
the introduction of "secondary" industries (mining and manufacturing) — 
which are called for because of the increased pressures from growing popula- 
tion and diminishing returns in the primary activities — as well as of the neces- 
sary social overhead facilities and services to support them; (4) a shift from 
a concentration on processing of farm and forest products and the simpler 
branches of textile, leather, and clothing industries, as well as on mining and 
mineral-reduction (usually based on outside' capital and enterprise) to more 
diversified industrialization based on internal industrial linkages and rising 
incomes; and, finally, (5) at an advanced stage of economic development, 
specialization in certain tertiary industries for export, including the export of 
capital, specialized personnel, and services to less advanced regions. 

8 See, for example, The Economic State of New England, Report of the Committee of New 
England of the National Planning Association, Arthur A. Bright, Jr., and George H. Ellis, 
directors of research and editors (New Haven: Yale University Press, 1954), pp. 3-23. The 
changing economic structure which has characterized the economic development of the New 
England states is employed as the frame of reference for the thoroughgoing study of the 
regional economy. 

9 August Losch, "The Nature of Economic Regions," Southern Economic Journal, Vol. 5 
(July 1938), pp. 71-78; Edgar M. Hoover, The Location of Economic Activity (New York: 
McGraw-Hill, 1948) , pp. 187-96. A valuable description of the stage theory is provided by 
Edgar M. Hoover and Joseph L. Fisher, "Research in Regional Economic Growth," Univer- 
sities-National Bureau Committee on Economic Research, Problems in the Study of 
Economic Growth (New York: National Bureau of Economic Research, July 1949), pp. 
180-88. 



60 / A Framework for Analysis 

An implication of the theory is that policies and activities of private groups 
and government can determine the rate at which the economy of a region will 
evolve from one stage into another; for example, in the building of social over- 
head facilities such as transport and in the efforts made to attract outside 
capital. 

The tie between the theory of development stages and the sector theory is 
apparent, and the one can be seen to be an extension of the other. Another 
extension of the sector theory takes the form of a widely held concept concern- 
ing the relationship between regional economic growth and the existence of 
"growth industries" within a region — that is, industries which have experienced 
a greater than (national) average increase in employment or in value of prod- 
uct or in earnings per worker over, say, the past decade or two. 10 The most 
common form in which this concept is expressed is in the presentation of argu- 
ments for consciously encouraging the expanding sectors — manufacturing and 
tertiary activities — as the way of achieving rapid economic growth. 11 A more 
specific concern with "growth industries" appears in the policy and activities 
of certain of the state and local "area development" agencies whose function 
it is to attract industries to a given region. Priority in effort tends to be given 
to those industries — a favorite is "electronics" — which have expanded relatively 
rapidly in the recent past. 

The sector theory has been found useful by a number of scholars as a frame 
of reference for the aggregation of data in the field of economic growth. Its 
focus on industrial and occupational structure, on changes in the pattern of 
demand (particularly with regard to income elasticity), and on relative changes 
in productivity has served to point up some significant economic-growth rela- 
tionships. 

LIMITATIONS OF THE EXPORT-BASE AND SECTOR THEORIES 

While both the export-base and the sector theories clarify important features 
of regional growth, their limitations are also apparent. The major limitations in 
both cases are (1) that they are partial in scope and overlook other equally 
significant aspects of regional economic growth and (2) that they deal with 
classifications which, while highly suggestive for general description, are too 
aggregate for analysis in depth. 

The export-base theory does not provide sufficient scope for the internal 
growth sequences — with the significant economic changes which these invariably 
involve — and therefore errs in the opposite direction from the sector theory. 
Residentiary employment does not adapt passively and automatically to the 
strategic "region-building" or "city-building" industries that have external 
trade relationships. Nor can all services be viewed as "dependent." Some 

10 See the discussion of "growth industries" in U. S. Department of Commerce Regional 
Trends in the United States Economy, A Supplement to the Survey of Current Business 
(Washington: Government Printing Office, 1951), pp. 5-7. 

"For a discussion of this point, see Harvey S. Perloff, "Interrelations of State Income and 
Industrial Structure," Review of Economics and Statistics, Vol. 39 (May 1957), pp. 162-63. 



Theories of Regional Economic Growth / 61 

provide the essential framework for the production of the export products and 
in a real sense are "lead" items from the standpoint of economic development. 
The interrelationships among the various kinds of activities within a region, 
including the complicated "multiplier" effects, are critical in regional economic 
growth. 12 

It is well to note also that the size of the region chosen for analysis inevitably 
has a great deal to do with the volume and importance of exports. The size 
question is particularly significant in the extent to which industry linkages 
come out as being "internal" or in the export category. And this, of course, 
greatly influences the view of the region's economic "strength" and its poten- 
tialities for growth. The problem arises out of the fact that in the application 
of the export-base theory, a given region or metropolitan community is analyzed 
in terms of its relationship to everything "outside." Douglass North has made 
the useful suggestion that, from the economic standpoint, the very definition 
of the region might well be in terms of a common export base as providing 
the unifying force for its development. 13 Such a view provides a useful frame 
of reference for a region in the earlier stages of development, or for one with 
quite limited factor endowment, but it runs into serious difficulty when applied 
to regions which are economically advanced and have a great variety of export 
industries whose markets range from limited areas to the entire nation or the 
entire world. To be examined comprehensively, economic growth in any signi- 
ficant area of the nation must be viewed in terms of the total complex of trade 
and other relationships, including interregional relationships and nodal-center 
to hinterland relationships, as well as in terms of ties to the national economy 
as a whole. 

The sector theory, unlike the export-base theory, brings "service" activities 
to the fore as important growth elements and is centrally concerned with in- 
ternal relationships generally; but, on the other side, it does not provide suffi- 
cient scope for the external relationships. It, too, hides from view the significant 
element of the functional ties between regions as well as the ties to the national 
economy (aside from the important factors of income elasticity of demand and 
labor productivity). Additional serious problems result from the grossness 
of the sectoral aggregates, that is, from the very great differences among the 
industries which are encompassed by the primary-secondary-tertiary classifica- 
tion with regard to their relationships to regional economic growth. (Still fur- 
ther difficulties attach to the two "extensions" of the sector theory with regard 
to development stages and "growth" industries. These are discussed in some 
detail in later sections.) 

12 For a discussion of these and other problems connected with the export-base theory, see 
Charles M. Tiebout, "Exports and Regional Economic Growth," Journal of Political Econ- 
omy, Vol. 64 (April 1956), pp. 160-64 (also the reply by North in the same issue, pp. 
165-68) : and the communications by Harris, Pfouts, and Tiebout on the base theory in the 
Journal of the American Institute of Planners, Vol. 24 (Fall 1958), pp. 233-46. 

13 "It is this that makes it economically unified and ties the fortunes of the area together. 
This tends to result in the interdependent development within the region of external eco- 
nomies and unified political effort for government assistance or political reform." North, 
op. cit., p. 257. 



62/^4 Framework for Analysis 

It would seem then that neither of these theories offers a really satisfactory 
framework for the type of broad-scale analysis of regional growth with which 
we are concerned. Each focuses attention upon some important aspects of the 
phenomena we study but leaves out other significant aspects of it. However, 
it is worth stressing that, though partial, these theories are highly useful for 
the insight they provide into the process of economic growth, and, as the 
reader will see, we shall try in what follows to show how they fit into the 
picture of economic growth in the United States. 

While we do not attempt in this volume to present a formal general theory of 
regional economic development, we shall first sketch out what we feel to be 
the significant features which such a theory should incorporate. In the empiri- 
cal analysis to follow later, we shall attempt as far as possible at this time to 
demonstrate how these features have worked themselves out in the past. Such 
an analysis, we believe, can make a useful contribution to a general theory of 
regional economic development. Possibly it is only upon a foundation of in- 
tensive empirical study that such a theory can be built. 



5 / Factors influencing the volume and 

composition of regional economic activity 



The fact that the secular and cyclical forces of the closely knit national economy 
dominate economic development in every part of the United States might suggest 
that we could best arrive at an understanding of regional growth by directly 
tracing to the regions the lines along which national economic growth are 
recorded. For example, between 1939 and 1954, employment for the nation as 
a whole increased by about 17%, or by 1.05% per year, and the real value of 
national output rose by about 92%, or by 4.5% per year. It is conceivable that 
these rates of growth in employment and output might have been the same for 
every region and for every industry sector — in fact, for every industry within 
every region. But, though this would greatly simplify the task of regional 
analysis, regions have not grown so, any more than have industries. There are 
wide regional departures from the national norm in both directions. To explain 
regional growth, therefore, we must explain these departures. 

As we see it, to understand economic growth within a particular region, it is 
necessary (1) to relate the region's development to developments in the nation 
as a whole; (2) to "weight" its growth in relative terms — that is, in terms of 
departure from the national norm; (3) to examine the characteristics of its 
growth pattern; (4) to evaluate its changing position with regard to its ability 
to hold and attract persons and industries; and, in general, (5) to study how it 
reacts to changes in the national "parameters" that influence supply and demand 
conditions for the major industries. 

Our analytical framework, therefore, provides for an examination of the 
following elements: 

1. The extent and character of national economic growth during the period 
that we are reviewing. 

2. The impact of change-initiating factors which are central to such growth, 
particularly (a) technology, (b) natural resources, (c) population and labor 
force (for example, changes in age structure which lead to more consumer- 
oriented living), (d) changes in taste, etc., which lead to changes in consumer 
demand, and (e) strategically important institutional changes, such as those 
flowing from governmental policy. In short, the analysis is concerned with the 

/ 63 



64 / A Framework for Analysis 

economic impact of the "external parameters" which are key parts in a vast 
gestalt of change through their effect on supply and demand. 

3. The relative extent to which the individual regions have shared in the 
national economic growth, and the shift in the relative position of the individual 
regions with regard to the key measures (such as employment within major 
industries). 

4. The major characteristics of the economic growth (or decline) patterns of 
the individual regions, particularly the extent to which such growth (or decline) 
is related to industry composition or to within-industry locational changes. 

5. The nature of the individual regions, particularly with regard to relative 
advantages in meeting the input and market requirements of the major indus- 
tries and with regard to the role of "agglomeration," as in the creation of 
external economies. 



National Forces 

The pattern of national economic growth and the forces behind it provide 
the logical starting point for regional analysis. In our economy, growth in 
total output is determined in large part by growth in the inputs of resources — 
labor, natural resources (e.g., land) , and reproducible capital — that go into the 
productive process and by technological and organizational changes that lead to 
greater output from a given input. 1 Particularly important here are improve- 
ments in knowledge, training, skills, etc., of the human resources. 

The key national data which need to be drawn on for the various aspects of 
regional analysis are: (1) growth in population, (2) growth in full-time equiva- 
lent workers employed, (3) change in output per man-year. These provide the 
figures for the change in total output per worker, as well as in the number of 
consumers. We then want to classify the employment and output (or income) 
data in terms of the major industries, so that the individual regional growth 
patterns can be compared with the national norm. 

Since so much of the analysis focuses on the developments within specific 
industries, the national change-initiating forces which are directly related to the 
growth of specific industries come to the forefront. Of these, productivity (on 
the supply side) and the income elasticities of demand (on the demand side) are 
particularly important. As population and per capita income grow, the desired 
composition of total output at given relative prices changes because of differ- 
ences in the income elasticities of demand for different commodities, resulting 
from varying physiological needs, changes in taste, introduction of new prod- 
ucts, and so on. Such data as we have at our disposal indicate, for example, that 
the income elasticity of demand for food is much less than one, while that for 
automobiles is much greater than one. Hence, as per capita income grows we 
would expect, other things being the same, that relatively more of society's total 

1 For an assessment of the contribution of these two forces to national growth, see Moses 
Abramovitz, Resource and Output Trends in the United States Since 1870, Occasional Paper 
No. 52 (New York: National Bureau of Economic Research, 1956), pp. 5-23. 



Volume and Composition of Regional Economic Activity / C5 

resources would be devoted to the production of automobiles and relatively less 
to the production of food. Likewise, technological change — particularly as it 
influences productivity — proceeds unevenly. 2 Because of this, the relative prices 
of different commodities change; consequently, the output of industries in which 
technological change has been the more rapid tends to increase at above-average 
rates. 

Since the forces which lead to national growth do not affect the demands for 
all commodities to the same degree, and since regions differ widely in the com- 
position of their output, such changes will stimulate marked growth in some 
regions while leaving other regions little affected. 

Technological change, it should be noted, not only works unevenly among 
industries at the national level, but may also work unevenly as among regions. 
With the change in production possibilities made available by changes in tech- 
nology, the pattern of regional advantage in production may change so that new 
capacity in an industry may be located in places other than those in which 
previously existing capacity was located. 

Changes in transportation costs, like changes in income elasticities and prod- 
uctivity, have a direct influence on the production and location decisions of 
individual industries and on their growth patterns. The impact of this factor 
varies greatly, of course; some industries are little influenced by changes in 
transportation costs. 

While industry shifts are the dominant factor in regional economic growth, 
some attention must also be devoted to population shifts. The migration of 
persons in response to income differentials, as well as non-job-oriented popula- 
tion movements, leads to significant changes in the size of markets and in the 
supply of labor in the various regions. 

These, then, are some of the major forces making for differential rates of 
change in the volume aspects of economic growth among the regions. They are 
so complex and so interrelated that we cannot directly translate the national 
parameters into regional growth effects. We find it helpful first to do an 
"accounting" job — specifically, to compare the growth patterns of the indi- 
vidual regions against the national norm — in order to analyze the relationship 
of the regional parts to the national whole. We do this mainly in terms of what 
might be called "share analysis" and "shift analysis." We then examine the 
national forces making for differential growth among regions specifically in 
terms of the changing input and market requirements of the major industries, 
mainly by way of what we call, for convenience, "input-output access" analysis. 
Here the locational shifts of industries and the reasons behind them are exam- 
ined in detail. (This is done largely in Part IV.) In this connection we also 
consider the "agglomeration" patterns, that is, the patterns characterizing the 
uneven geographic distribution of population and the various industries. These 
are important in defining markets and, because of the economies and disecono- 
mies which they introduce, in shaping the cost differentials for economic activ- 

2 For estimates of the extent of technological change in different industries see John W. 
Kendrick, Productivity Trends: Capital and Labor, Occasional Paper No. 53 (New York: 
National Bureau of Economic Research, 1956). 



66 / A Framework for Analysis 

ities in various parts of the country. They are important also because, within 
broad geographical regions, they lead to local variations in demand and supply. 
These patterns, together with other specifically regional factors, such as varia- 
tions in natural resources endowment, climate, amenities, and the like, condition 
the way in which a specific region can react to a given national change. 

We shall take up these various accounting and analytical elements in turn, 
starting with a discussion of "shares." 



Shares of the Whole 

It is necessary as a starting point to establish the relationship of the (re- 
gional) parts to the (U. S.) whole 3 and to each other. This is so whether the 
focus is on all parts of the country (as is the case here) or on a single region. 

The economy of a region can be usefully characterized in terms of the shares 
of the U. S. total economy which it contains or contributes, and important 
insights into economic growth can be obtained by tracing the changes in these 
shares. The totals and shares to be measured will, of course, depend on the 
central interest of the analysis; for purposes of growth analysis, interest focuses 
on the growth indexes — population, income, etc. — and on the industrial struc- 
ture, particularly as reflected in employment and in production by industrial 
sectors. 

Each part of the country plays a more or less specialized role in the total; it 
produces and consumes a qualitatively unique share. One part may have basic- 
ally a "breadbasket" function; another may be a "heavy manufacturing" zone; 
and a third may have a "playground" function (as is the case for large sections 
of Nevada and Florida). An analysis of this division of functions is complicated 
by the large size of the nation — both in area and in terms of markets (so that 
most products are produced in more than one center) — and by the gradualness 
of the developmental sequences. And, of course, there are certain ubiquitous 
functions and activities, such as retailing and local services, which need to be 
accounted for. 

Thus, in addition to recording the quantitative shares, it is necessary to pro- 
vide qualitative depth by highlighting the unique or specialized elements as com- 
pared to the ubiquitous elements in the economic structure of each region. 4 
Equally important from the qualitative standpoint are certain relationships 
among the shares which, in an important sense, characterize the growth patterns 

3 The "whole" which is the logical focus of analysis is itself an analytical variable and 
depends on the nature of the markets and trade patterns which dominate. Thus, study of 
the U. S. regions in the 18th century would have to relate the regions to the "Atlantic 
Community"; in the 20th century, the national economy logically serves as the larger frame- 
work for regional analysis. 

4 A number of valuable measures of industrial concentration within a given area as com- 
pared to the national average have been developed in recent decades. The "coefficient of 
localization," as developed by P. Sargent Florence has been widely employed. See his 
treatment of this measure in National Resources Planning Board, Industrial Location and 



Volume and Composition of Regional Economic Activity / 67 

of the various regions. An obvious example is the comparison of a given 
region's share of the total national population and its share of total income over 
a period of time. Another example is the connection between a region's share 
of total employment in the resources industries and its share of the output value 
of these extractive industries. Thus the following comparison between the Far 
West and the Southeast provides an important lead for analysis: In 1870, the 
Far West had 17% of the total national labor force in resource activities and an 
exactly equal share of the total value of extracted resources; in 1950, its shares 
were 38% of the labor force and 54% of the value. The comparable figures for 
the Southeast in 1870 were 40% of the total resources labor force and 25% 
of the value of extracted resources; and in 1950 they were 38% and 20% of the 
national totals, respectively. By such a comparison — particularly when the 
resource industries are broken down into their components (agriculture, fishing, 
forestry, and various kinds of mining) and the "greater regions" are disaggre- 
gated into state units — changes in volume of activities and in relative gains in 
productivity are brought to the forefront. 

The regions must be seen not only in relation to the national totals, but also 
in relation to each other. It is useful for purposes of analysis to think of the 
various regions as in competition for shares of the national totals. Thus, it is 
not enough to know that a particular region currently has a smaller share of the 
total national employment and production of a given industry, for example steel, 
as compared to a generation ago ; it is important to know also about the regions 
that gained in shares over the same period. Was their employment gain directly 
related to their population gain, or was it larger than their population gain? The 
implications are very different in the two cases. Did they make a productivity 
gain or only a volume gain? Again, the implications are quite different. Other 
types of analyses obviously need to be brought to bear on such questions — and 
these will be discussed in the sections that follow — but here we want to establish 
the value of share analysis as providing an essential background of information 
for the understanding of regional economic growth in an appropriate context. 

In general, it can be said of share analysis that: (1) it helps to define the 
regions by relating them quantitatively, and to some extent qualitatively, to the 
national totals and to each other; (2) it usefully characterizes the economic 
structure of the various regions, again in relative terms; (3) it highlights the 
extent to which various aspects of growth are uniform across the entire nation, 
or large parts of it, and the extent to which they involve shifts from some 
regions to others; and (4) it helps to characterize the relatively stable and the 
less stable elements in the national economic structure and in the structures of 
the individual regions. 

National Resources (Washington: Government Printing Office, 1943), pp. 63-84 and 105-24. 
Also the discussion of a "location quotient" by George Hildebrand and Arthur Mace, Jr., 
"The Employment Multiplier in an Expanding Industrial Market: Los Angeles County, 
1940-47," Review of Economics and Statistics, Vol. 32 (August 1950), pp. 241-49; and the 
discussion of the subject by Otis Dudley Duncan and Associates, Metropolis and Region 
(Baltimore: The Johns Hopkins Press, 1960). 



68/^4 Framework for Analysis 



Patterns of Industrial Growth Within Regions 

The type of economic change that occurs within a given region depends 
mainly on two things: (a) the economic activities that are subject to the most 
significant nation-wide changes and (b) the nature of the given region. In 
broad terms, then, what has to be done is to trace the regional variations on the 
major national themes. Let us consider first the influence on regional growth of 
nation-wide changes in industry growth patterns. 

It is often assumed that those regions will grow that have the greatest relative 
net advantages in attracting the more rapidly growing industries. In an exami- 
nation of the important sector of manufacturing, however, we found no positive 
correlation between increases in the proportion of workers within the "growth 
industries" and relative increases in economic activity in general among the 
states. 

We examined here the relationship between growth in income (total and per 
capita) and manufacturing employment for the 1939-1954 period. The aspect of 
structure considered was the percentage of total manufacturing employment in 
those industries which grew most rapidly for the nation as a whole during these 
years. The industries selected — nine in all — were all industries in which the 
relative increases in employment between 1939 and 1954 exceeded the average 
for all manufacturing; it is worth noting that they accounted for about half the 
total employment in manufacturing in 1950. 5 

The rank correlation coefficients for both income variables (total and per 
capita) and manufacturing employment with the percentage of manufacturing 
employment in these industries were all negative; that is, states which had the 
largest percentage of manufacturing employment in these more rapidly growing 
industries tended to have the smallest increases in income, both total and per 
capita, and in total manufacturing employment. There was a positive but small 
rank correlation between population increase and percentage of manufacturing 
employment in these eight industries by states. 

The reason for these negative results can be readily seen by examining Figure 
18. The states with the highest percentage of manufacturing employment in the 
rapidly growing industries were generally in the Northeast, where growth was 
generally less rapid during 1939-1954, while those with the lower proportions 
in these industries were in the Southeast, where growth in income and total 
manufacturing employment was generally among the most rapid in relative 
terms over these years. 

Essentially similar results were obtained by comparing the growth measures 
listed above with the percentage of the increment in manufacturing employment 

5 The selection was made by examining trends for the nation as a whole for the 1939-1954 
period using Department of Commerce data on wage and salary payments and full-time 
equivalent employees. The industries were the following: (1) transportation equipment; 
(2) chemicals; (3) electrical machinery; (4) petroleum and coal products; (5) paper 
products; (6) machinery except electrical; (7) metals; (8) instruments; and (9) miscel- 
laneous manufacturing. Employment data for states were obtained from the 1950 population 
census. Also see Table 130. 



Volume and Composition of Regional Economic Activity / 69 




Figure 18. Percentage of Manufacturing Employment in the More Rapidly 
Growing Industries, by State, 1950. 

Source: U. S. Bureau of the Census, U. S. Census of Population 1950, 
Vol. II, State Table 80. 



during 1939-1954 that was in these same nine industries, as revealed by data 
from the manufacturing censuses of these two years. (See Figure 19.) 

The implications of these findings are significant for the understanding of 
regional economic growth. There are two elements here that seem particularly 
important. One is that a region may grow through getting an increasingly 
greater proportion of an industry that as a whole is declining nationally. This 
would be the case when the remaining portions of a declining industry all cluster 
into one or two regions, which gain in volume with regard to this industry while 
all others are losing out. Or a region may grow because it contains the growing 
parts of a generally declining industry. Since industry categories employed for 
economic analysis always contain sub-categories (this is highlighted by the 
Census designation of 1-digit, 2-digit, 3-digit and 4-digit categories), 6 some of 
the sub-categories will normally be expanding even when the average for the 
whole group shows a decline; this is the case, for example, with the growing 
cattle industry in the generally declining over-all industry of agriculture. Con- 
versely, the textile industry is a declining sub-category of the generally growing 
manufacturing category. When these various possibilities are combined, it can 

"The higher the number of digits, the more detailed the classification. Thus "mining" 
is a 1-digit category, (a) Petroleum and natural gas, (b) Nonmetallic mining, (c) Metallic 
mining, and (d) Coal mining are the 2-digit subcategories. The metallic mining category is 
further disaggregated into 3-digit classes, including such industries as iron ore, gold and 
silver, copper, and so on. 



70 / A Framework for Analysis 




> 86.4% 
123 68.0-86.4% 
fcsH 46.0-67.9% 
fv^ < 46.0% 
d = disclosure problem 



Figure 19. Increase in Manufacturing Employment in the More Rapidly 
Growing Industries, as a Percentage of the Increase in All Manufacturing 
Employment, 1939-54. 

Source: Appendix Table K. 



be seen that a given region may contain, at one extreme, an increasing share of 
a growing sub-category of a nationally-increasing industry or, at the opposite 
extreme, a declining share of a declining sub-category of a nationally declining 
industry, with many combinations in between. 7 



Proportionality and Differential Shifts 

Regional economic growth can be understood only when all of this can be 
clarified, measured, and explained. The statistical technique that we have 
employed for this purpose is an elaboration of the "shift" technique described in 
Chapter 3. It is based upon the fact that the shifts in total employment (or in 
other important economic growth components) observed among the states and 
regions in relation to the national average are generated by two distinct types of 

7 Emphasis on the fact that regional economic growth is not simply a matter of attracting 
the so-called rapid-growth industries should in no way diminish the significance for economic 
expansion of the presence of such industries. Clearly, a growing industry within a region 
is a stimulus to over-all growth. This is so evident that it does not require emphasis. The 
other side does require emphasis; namely, that a region can enjoy a substantial amount of 
over-all economic growth by absorbing a larger and larger share of a declining industry or 
by attracting the growing parts of an industry which is declining on the average. This, as 
we will note later, will be particularly effective if the industries are of the type which tend 
to generate a strong multiplier effect. 



Volume and Composition of Regional Economic Activity / 71 

phenomena. We call the resultant employment shifts "proportionality employ- 
ment shifts" and "differential employment shifts." 8 

The net differential shift in employment arises out of the fact that some regions 
are expanding in certain employment sectors (say, in the steel industry or 
personal services) more rapidly than other regions. The regions that show net 
upward differential shifts, as we shall see, are those in which over-all access to 
basic inputs or to markets has been improved relative to other regions engaged 
in the same activity; they have gained because of their greater looational 
advantages for the operations of given activities or industries. 

Proportionality shifts are distinct from differential shifts. They arise out of 
the fact that, nation-wide, some of the employment sectors or industries expand 
more rapidly than others. (That is, here the "growth industry" effect is re- 
corded.) As a consequence, those regions that tend to specialize in the slow- 
growth sectors (e.g., agriculture and mining) show net downward proportion- 
ality shifts in total employment, while those that tend to specialize in the 
rapid-growth sectors show net upward proportionality shifts. 

There are two ways by which the net proportionality shift in total employment 
can be computed. One way would be to use a complex weighted average based 
upon the degree of specialization in each sector by each state and the rate of 
employment growth for each sector. Obviously this is a very time-consuming 
method. The second way is based on the premise that the net differential shift 
in total employment and the net proportionality shift in total employment add 
up to the net total employment shift. Thus we can simply subtract the net 
differential shift in total employment from the net total employment shift to 
obtain the net proportionality shift in total employment. 9 

Because of the importance of this analytical tool in the discussion of regional 

8 This involves a standardization procedure of the type employed when changes in rates 
and composition are both of interest. See Frank A. Hanna, "Analysis of Interstate Income 
Differentials: Theory and Practice," Regional Income, National Bureau of Economic 
Research, Studies in Income and Wealth, Vol. 21 (Princeton: Princeton University Press, 
1957), pp. 113-60; also the discussion by Denison, Williams, and Borts, pp. 161-93. 

9 The following may help clarify this point. Let : 

Eij = employment in the i-th industry and j-th state in the initial time period 

E*ij = the same in the terminal period 

Ei. — national employment in the i-th industry = SjEij 

E.j — total state employment — S ; Ey 

E.. — total national employment in all industries = Sj,j Eij 

Then, the total shift for a state is: 

(1) S t = E*.j - (E*../E..) E.j 
The differential shift for the state is: 

(2) S d = Si [E* u - (E*,./E,.) Eij] 
And, 

S t - S d = TE*.j - (E*../E..) E.j] - S, [E* u - (E*i./Ei.) Eij] 
= 2. (E*i./Ei.) Eij - (E*../E..) Si Eij 

(3) = Si [ (E*,./Ei.) - (E*../E..) ] Eij 

But, (3) is simply the proportionality shift, S p , a weighted average of the excess of 
national industry growth rates over that for all industry nationally, where the weights for a 
given state are its total industry employment in the initial period. 



-J_ 


139,333 


+ 


3,163 


_l_ 


77,708 


_l- 


435,374 


_j_ 


74,799 


+ 


12,267 


+ 


91,303 


+ 


28,860 


_!_ 


45,615 


+ 


224,766 


+ 1,133,188 



72/^4 Framework for Analysis 

economic growth which follows, it may be well to provide an actual example 
here. Between 1939 and 1954, California had an increase in total employment 
of 2,116,282 workers. If its employment had grown at the same percentage rate 
as for the country as a whole over these years, its increase would have been only 
some 812,477 workers. Thus, California can be said to have had a net upward 
shift in employment between 1939 and 1954 of 1,303,805 — the difference 
between the two figures. 

If we apply the same technique to each industry, instead of to only the total, 
we get the following result: 

Agriculture 

Mining 

Construction 

Manufacturing 

Transportation and Public Utilities 

Wholesale Trade 

Retail Trade 

Finance, Insurance, and Real Estate 

Service and Miscellaneous 

Government 

Total 



In other words, every major industry in California had a greater increase in 
employment than it would have had if it had grown at the national rate; in 
manufacturing, the increase was some 435,000 more than "expected." However, 
the total — 1,133,188 — is less than the total net shift in employment indicated 
earlier; less by 170,617. The reason is that, not only did each industry in Cali- 
fornia grow more than the national average for the industry, but California's 
industry mix or composition was such that the number of workers employed in 
"growth industries" exceeded the national average. This effect is reflected in 
the 170,617 difference noted above, or what we call the "proportionality effect." 

The California situation of growth in every department is, of course, only 
one type of situation observed. Other states have had a different experience. 
Table 13 shows the differential shifts for Texas, Pennsylvania, and Georgia in 
addition to California. In deriving the figures presented here, we first computed 
the net upward or downward shift in each of the various employment sectors. 
The net differential upward or downward shift in total employment for each 
state is the algebraic sum of these shifts in the component sectors. 

Again we note that California had a net upward differential shift of 1,133,188 
wage jobs based upon upward shifts in every component sector. Texas also 
shows a net upward shift, but the drop in agricultural employment here dampens 
the total performance. Pennsylvania, by contrast, shows net downward shifts in 
virtually every sector except agriculture and wholesale trade. Georgia's total net 
shift is downward in spite of upward shifts for most of the component sectors. 
The downward shift in agriculture is so large that it overwhelms the gains in the 



Volume and Composition of Regional Economic Activity / 73 
Table 13. An illustration of the differential shift for selected states, 1939-1954 



Major 












industrial sectors 


California 


Texas 


Pennsylvania 


Georgia 


Agriculture 


+ 


139,333 


- 64,896 


+ 4,734 


-171,044 


Mining 


+ 


3,163 


+ 80,347 


- 86,653 


+ 772 


Construction 


+ 


77,708 


- 6,911 


- 3,742 


- 12,471 


Manufacturing 


+ 


435,374 


+ 142,854 


-240,017 


+ 10,760 


Transportation and Public 












Utilities 


+ 


74,799 


+ 35,060 


- 29,609 


+ 4,710 


Wholesale Trade 


+ 


12,267 


+ 24,166 


+ 3,797 


+ 14,285 


Retail Trade 


+ 


91,303 


+ 54,224 


- 64,242 


+ 8,232 


Finance, Insurance, and Real 












Estate 


+ 


28,860 


+ 33,328 


- 4,483 


+ 9,561 


Service and Miscellaneous 


+ 


45,615 


+ 33,436 


- 22,999 


- 7,252 


Government 


+ 


224,766 


+ 77,261 


- 102,696 


+ 26,736 


Total differential net shift 












in employment 


+ 1,133,188 


+ 408,869 


-545,910 


-115,711 



Source: Appendix Table H. 



other sectors. (An explanation of the behavior of these states and others is 
offered later in the text.) 

Suppose that, by an extraordinary coincidence, none of the component sectors 
shown in Table 13 had any net upward or downward shift (that is, no change 
in position as compared to all other areas of the nation). It would follow that 
the differential net shift for each state would be zero. It would not follow, how- 
ever, that there would be no shift in total employment for these same states 
because the proportionality effect remains as a source of shifting employment 
shares. 

Table 14 sets down the total net shift in employment for the 1939-1954 
period and the total differential net shift from Table 13 for these four states. In 
no case does the differential net shift match the shift in total employment. It 
either fails to exhaust the total or is larger than the total. The balance or deficit 
in each case is attributable to the proportionality effect, which arises out of the 
fact that, nationwide, some of the employment sectors themselves expand more 
rapidly than others. 

While it is only a limited illustration, Table 14 reveals that several kinds of 
situations can develop. California experiences an upward shift in total employ- 
ment because both the differential and proportionality effects contributed to it. 
Texas experiences an upward shift in total employment but only because a very 
strong upward differential shift is sufficient to offset a significant downward 
proportionality shift based upon this state's relative specialization in mining and 
agriculture. Pennsylvania reverses the experience of Texas. This state experi- 
ences a downward employment shift because a strong downward differential 



74/^4 Framework for Analysis 



Table 14. Illustration of total employment shifts derived from a combination 
of the differential effect and the proportionality effect for selected states, 1939- 
1954 



Total employment shift California Texas Pennsylvania Georgia 

Net shift in total 

employment +1,303,805 +112,286 "276,399 -273,931 

Net differential (within industry) 

shift in total employment +1,133,188 +408,869 -545,910 -115,711 

Net proportionality (industry growth) 

shift in total employment + 170,617 -296,583 4 269,511 -158,220 

Source: Appendix Table H. 



shift is sufficient to offset a significant upward proportionality shift based upon 
relative specialization in manufacturing, trade, and services. 10 In Georgia, the 
differential effect and the proportionality effect reinforce each other in gen- 
erating downward shifts in total employment. 

This statistical device provides a useful framework for analysis. Sources of 
change in the regional structure of the economy are divided into two classes. 
Each requires a distinctly different type of analysis. The differential effect arises 
out of the fact that some regions gain, over time, a differential advantage (vis-a- 
vis other regions) in their access to important markets and inputs for each of 
one or more specific activities. An understanding of this effect involves an under- 
standing of regional input-output relationships for specific activities and sources 
of their changing form. One must become deeply involved in location analysis. 
The proportionality effect arises out of the fact that the various regions start 
with a different industry mix or composition — that some regions claim a larger 
(or smaller) proportion of the nation's rapid-growth (or slow-growth) indus- 
tries. An understanding of this effect involves one in an analysis of the changing 
total demand and supply relationships for the individual industries, including 
such elements as income elasticity of demand, changing tastes, technological 
developments, as well as an analysis of the special advantages of a given region 
for such rapid-growth or slow-growth industries. 

An examination of regional change must clearly take into account both types 
of phenomena. 



10 New Jersey, in contrast, displays an upward employment shift for the 1939-54 period 
because the downward differential shift is not strong enough to offset a significant upward 
proportionality shift based upon its relative specialization in rapid-growth sectors. 




Location factors 



Location theory 1 suggests that growth in a given area's volume of economic 
activities is directly related to two factors: its access at competitive costs to the 
inputs of production and its access at competitive costs to markets for the 
outputs of this production. The quantity and quality of a region's resources are 
therefore significant for growth; and, because so much of manufacturing 
involves the fabricating of processed materials, with many stages of value 
added, the availability of intermediate inputs may be equally or more signifi- 
cant. The size of the regional market and the proximity (in terms of transport 
costs) of national markets are also important considerations. The regions differ 
widely in these attributes. To understand the dynamic elements through which 
these relative regional advantages and disadvantages come into play, it is neces- 
sary to appreciate the principles behind location decisions in our economy. 



Location Principles 

The general principles that govern the location of economic activity are 
essentially the same as those that govern all of man's efforts to make the most 
of his economic resources. On the production side, expenditures for inputs 
reflect an effort to produce the product-mix buyers want most with the smallest 
possible commitment of economic resources. On the consumption side, expendi- 

a The discussion of location factors in regional economic growth is largely based on 
materials prepared by Edgar S. Dunn. These factors will be taken up in some detail at a 
later point. 

For valuable discussions of location theory, see Alfred Weber, Theory of the Location of 
Industries, edited by C. J. Friedrich (Chicago: University of Chicago Press, 1920); E. M. 
Hoover, Jr., Location Theory and the Shoe and Leather Industries (Cambridge: Harvard 
University Press, 1937) and The Location of Economic Activity (New York: McGraw-Hill, 
1948) ; Bertil Ohlin, Interregional and International Trade (Cambridge: Harvard Uni- 
versity Press, 1952) ; William H. Dean, The Theory of the Geographic Location of 
Economic Activities, (Ann Arbor: Edwards Bros., 1938) ; National Resources Planning 
Board, Industrial Location and National Resources (Washington, D.C.: Government Print- 
ing Office, 1943) ; Walter Isard, Location and Space-Economy (New York: Technology 
Press of Massachusetts Institute of Technology and John Wiley & Sons, 1956) . 

/ 75 



76/^4 Framework for Analysis 

tures for products reflect the desire to acquire a market basket of items that 
will provide the maximum satisfaction possible within the restraints of in- 
dividual incomes. The complex set of decisions that constitute man's economiz- 
ing effort are made largely within the framework of the market mechanism. 

Both facets of location are discussed here: the location of productive activity 
and location of consumption activity. 

LOCATION OF PRODUCTIVE ACTIVITY 2 

The businessman's principal concern in picking a production site is to select 
one that will enable him to operate at the most profitable level of output. Pro- 
duction costs and revenues vary from one site to another, depending upon 
variations in access to the basic inputs and markets that are significant for 
specific activities. Sometimes input access is dominant in determining the loca- 
tion of an activity, and sometimes market access ; sometimes neither is dominant, 
with the result that an intermediate or indeterminate location develops. This has 
led writers in the field to characterize production activity as being market- 
oriented, material-oriented, or oriented to intermediate sites. 3 

This basic grouping of industries with similar location characteristics is a 
helpful descriptive and analytical device. However, we have found it useful to 
make two changes in the classification. First, the term "material orientation" 
is not altogether an appropriate one. It would be better to speak more generally 
of orientation to input requirements, for there is often orientation to inputs 
that are distinctly non-material in character. This is true particularly of labor 
services, which constitute the principal input for most businesses. Second, the 
classification scheme can usefully be enlarged, as follows: 

I. Activities that maximize their total access by orienting primarily to 
inputs 

A. Resource inputs 

2 Here we are using production in its broadest sense, including all extractive, processing, 
trade and service occupations that contribute to the economic welfare of the populace. 

^Market orientation may be of different types. Many market-oriented activities — for 
example, the production of bakery goods and consumer services — are oriented specifically to 
final markets. Many others, such as machine shops, are oriented to intermediate markets. 
A few industries, such as production of fertilizers, tin cans, etc., are oriented essentially to 
resource activities as markets. 

The same diversity can be observed in input orientation. Many input-oriented activities, 
such as coal-mining, are oriented to resource inputs. Others, such as paper-manufacturing, 
are oriented to intermediate inputs. And a few, such as scrap and waste collection, are 
oriented to inputs primarily generated at market sites. 

Many activities, such as wholesaling and automobile production, are intermediate to both 
markets and inputs. Some activities, such as petroleum-refining and some petrochemicals 
industries, have mixed orientation patterns. Under one combination of markets and inputs, 
they tend to be market-oriented; another combination may lead to input-orientation. It is 
important to realize that any given production activity might have a different location 
orientation at the same time in different regions. 

Some activities, such as hosiery production or some branches of electronics, do not have 
any strong orientation pattern. Such industries may be footloose in varying degrees. 



Location Factors / 11 

B. Intermediate inputs 

C. Market inputs 

II. Activities that maximize their total access by orienting primarily to 
markets 

A. Final markets 

B. Intermediate markets 

C. Resource markets 

III. Activities that maximize their total access by selecting locations inter- 
mediate to inputs and markets 

IV. Mixed location patterns 

V. Production activity without specific orientation (footloose) 



LOCATION FACTORS 

The costs attendant upon the assembly of inputs and the distribution of out- 
puts can vary a great deal from one region to another. And these variations, 
together with interregional price differences, are the main determinants of the 
location patterns. Thus transfer relations with input sources and markets are a 
major factor in explaining the location of various industries. 

There are two special situations, however, in which transfer relations are 
unimportant in determining location. In one, transfer costs may be significant, 
but the transfer alternatives offer no choice. In the other, transfer alternatives 
may offer many choices, but transfer costs are insignificant. 

The first of these situations is illustrated by the immobile input that is 
indispensable and for which there is only a single source and/or no substitutes. 
In such a case the production activity utilizing the input must turn to the single 
source of supply, and transfer costs become irrelevant in the decision. 

The second type of situation often develops in relation to a product or service 
for which transfer costs are very small relative to the value of the product or 
service. A hosiery mill's total transport cost on material inputs and products, 
for example, is less than 1% of its operating costs; and some branches of the 
electronics industry assemble their materials for less than one-fourth of 1% 
of their total costs. Obviously, in these cases, a substantial percentage differ- 
ence in transfer costs could exist between regions without changing the operat- 
ing costs of the enterprise by as much as one-fourth of 1%. For such enter- 
prises, the location choices are almost unlimited and the final choice is to a 
considerable degree objectively indeterminate. 

Except in these two types of situations, in which location is either fixed or 
indeterminate, transfer costs can exercise influence on an industry's choice of 
location. This influence can be operative even in cases where inputs or outputs 
are immobile if there are substitute sources for the same input or substitute 
markets for the same output. For example, mining activities are inevitably 
oriented to the immobile natural site of the resource. However, where there are 
several alternative sites, that one will be selected that gives the mining firm the 
greatest total access to the essential resource input and its potential markets. 



78/^4 Framework for Analysis 

In situations of this type, the existence of transfer alternatives is a major factor 
in determining the order in which resource sites may be brought into use and 
the intensity with which they may be exploited. 4 

For the vast majority of enterprises, production inputs and outputs are mobile 
and substitutable, and transfer relations are important. These characteristics 
make every possible site in the economic landscape a potential production site. 
However, such activities must resolve the transfer costs associated with inputs 
and outputs favorably. As a result, economies with reference to sources and 
markets help make specific the location pattern and type of orientation for each 
enterprise. This potential flexibility of space arrangements presents the oppor- 
tunity for a host of factors to influence access through transfer costs and hence 
to influence the location decision. 

The ratio of transfer costs on inputs as compared with outputs is of major 
importance among these influences. In general, a higher rate of transfer costs on 
inputs relative to outputs over the same distance tends to make the production 
activity input- oriented. The location pattern of pulp and paper-manufacturing 
is a good example of this influence. A reverse ratio of transfer costs favors a 
market-oriented production site. Brick and concrete products are good examples. 
These transfer costs depend, in turn, upon a number of factors. The relative 
weight and bulk of the products and the inputs are important. Weight losses 
(as in beet sugar processing) or weight gains (as in the production of soft 
drinks) in the production process and differences in transport rates per ton/ 
mile on inputs and outputs strongly influence transfer ratios. The time factor in 
transfer imposed by perishability or special service requirements is important. 5 

Another set of factors influencing the location of an enterprise is the spatial 
character of its supply sources and markets. Actually orientation in terms of 
strict juxtaposition to markets or supply sources can occur only if the market 
or the supply source is essentially pointform, that is, if it exists only at a single 

"This process involves three elements. (1) The direct transfer costs in distributing the 
output of the mine which will give some mine locations an advantage over others. (2) The 
opportunity costs that arise out of the fact that natural resource deposits vary in quality. 
(3) Rivalry costs associated with the competition of other firms for scarce resource deposits 
and for the market. The firm will attempt to minimize the combined costs from all three 
sources. It is important to note that the rivalry costs have, themselves, a significant space 
dimension derived from the management decisions of the rival firms. To focus upon the 
influence of space on the direct transfer costs of one firm would ignore the important in- 
direct influence of space imposed by a rival's choice. 

5 In general the ratio of transport costs on inputs and outputs will result in orientation 
either to markets or input sources. Intermediate sources face two serious handicaps. 1) 
An intermediate source will require an additional handling charge because it introduces 
a transshipment. 2) The graduated rate structure that makes short hauls relatively more 
expensive than long hauls tends to make prohibitive the combined hauls necessitated by an 
intermediate site. There are two circumstances, however, that may offset this handicap in 
specific cases. 1) Breaks in transport (i.e., between rail and water) between the input 
source and market require a transshipment anyway and often present a feasible inter- 
mediate location point. This is a factor that adds considerably to the economic importance 
of port cities. Sugar refining in Atlantic seaboard ports forms a good example of this 
influence. 2) Peculiarities in transport rate structure sometimes make intermediate points 
favorable sites from the point of view of transfer costs. Sometimes this result is accidental 



Location Factors / 79 

point in space. Markets are seldom pointform for those production enterprises 
that are oriented to markets associated with final demand. 6 People do not exist 
at one or even a few points in space; they are areally distributed. Thus, produc- 
tion activities directly serving their wants almost always find their markets 
areally distributed. Market orientation in such a case is only relative and 
involves determining an approximate transport center for an areally distributed 
set of market transfers. Retailers, wholesalers, food processors, and personal 
service enterprises of all kinds exemplify this situation. 

A similar situation often arises with reference to resource supply sources — 
particularly if the production process is oriented to agricultural inputs. Pulp 
mills, cotton gins, flour mills, meat-processing plants, and milk-processing plants 
all must orient their production with reference to areally distributed supply 
sources. Input orientation involves determining an approximate transport center 
for an areally distributed set of input transfers. 

The general influence of market and supply areas is to increase the impor- 
tance of intermediate sites. Where a production process has a supply area but 
an essentially pointform market outlet (such as a single firm at one point in 
space), this circumstance might tend to strengthen the pull of the market. Where 
a production process has a market area but an essentially pointform supply 
source, this might tend to strengthen the pull of the important input. Produc- 
tion enterprises that are oriented primarily to intermediate inputs and outputs 
would have the greatest chance for having pointform markets and supply 
sources. There would, therefore, be a greater chance for final orientation to 
markets or inputs in these cases. 7 

The internal scale economies of an individual enterprise are likewise of signifi- 
cance for location patterns. By economies of scale, we refer to the fact that 
technological production characteristics determine whether, to produce effi- 
ciently, an enterprise must produce on a very large, medium, or small scale. 
Scale economies leading to firms of large size have the effect of reducing the 
number of feasible input sources (in the case of input orientation) or markets 
(in the case of market orientation). If a firm is market-oriented, it must have a 
market large enough to allow for the operation of at least one minimum-sized 

or it results from a peculiar combination of "block rates." More often it is deliberate 
and is a result of an "in transit" privilege in which the carrier grants the "through rate" 
to an intermediate producer even though there is a transshipment involved. The location 
of flour mills and soybean mills in the Midwest have evidently been influenced by this 
factor. 

"Unless final demand presents itself in the form of a governmental unit or an export 
demand. 

'The location character of a production enterprise is also significantly influenced by the 
number of inputs and/or outputs that are characteristic of its production function in a way 
similar to the influence of the spatial character of supply sources and markets just discussed. 
Even in a case where all production inputs are pointform, multiple production inputs will 
tend, as does a supply area, to encourage location at an intermediate location. There are 
exceptions (1) when the production inputs are all coincident in space and (2) when one 
of the inputs is locationally dominant (i.e., its location weight exceeds the combined 
weight of all other inputs and products). The same generalizations apply to multiple 
products and their influence upon location. 



80 / A Framework for Analysis 

plant. In the extreme case where one firm (or a few firms) can efficiently pro- 
duce an output adequate for the national market, production will inevitably be 
considerably concentrated in this field. One region may have excellent transfer 
relations with reference to inputs and markets and therefore represent an 
excellent location point for producing a particular product. If, however, a 
producer in another region (likewise with good transfer relations) has pre- 
empted the total market, this excellence of site will avail nothing. At the other 
extreme, small-scale efficiency relative to market size usually results in wide 
dispersion of activities of this type. 8 

Another set of factors that exert an influence upon location patterns are 
external scale economies. Once a firm has achieved its minimum efficient size in 
terms of its internal technological structure, it may enjoy additional cost savings 
originating from sources external to itself. 

One source of these savings is an increase in efficiency within an industry as 
a whole. The growth of an industry (such as the automobile industry) may 
enable a specialized supplier to this industry (such as an ignition manufacturer) 
to achieve internal scale economies that result in lower costs to the firms in the 
base industry. Growth in the number of firms in an industry may permit the 
organization of specialized services devoted to that industry (such as the re- 
moval or processing of waste materials). Joint research, organized markets for 
finished products, specialized brokers, and specialized machinery producers are 
all examples of the kinds of economies that may be involved. 

Another source of external cost savings is the increase in efficiency that results 
when a group of unlike enterprises grow together. Organized raw-material 
markets, more favorable freight rates, the generation of a large body of labor 
with specialized skill and training, the availability of specialized services for 
maintaining business machines, and important community services such as 
power and water are all examples of important external economies of this type. 

Not all external economies available to firms in an industry require that the 
firms be located in an area of concentration. Such things as the advantages of 
organized exchange, improvements in machinery resulting from the growth of 
an industry, and fundamental or applied research are available to all firms 
irrespective of their location. Many of the others, however, are available only to 
firms concentrated in an area. Thus, external economies may be mobile or 
immobile. 9 Where the economies result from areal concentration, the term 
"agglomeration economies" is applied. 

Production enterprises in the U. S. exhibit great variety in this respect. A plant of 
minimal efficient scale producing typewriters can produce up to 30% of the national 
industry capacity. By contrast, a cement mill of minimal efficient scale produces less than 
1% of the national capacity. For an additional classification of industries in this regard, 
see J. S. Bain, "Economies of Scale, Concentration, and the Condition of Entry in Twenty 
Manufacturing Industries," The American Economic Review, Vol. 44 (March, 1954), 
pp. 15-39. 

9 This distinction has been underlined by E. A. G. Robinson, The Structure of Competi- 
tive Industry (New York: Pitman Publishing Corp., 1959), p. 142. See also John A. Guthrie, 
"Economies of Scale and Regional Development," Papers and Proceedings of the Regional 
Science Association, Vol. 1 (1955), pp. J1-J10. 



Location Factors / 81 



Agglomeration 



It is evident that the restraints imposed by space on the movements of inputs and 
products importantly influence decisions on the location of productive activity. 
Thus a region's position in the spatial framework of the national economy 
directly affects its economic growth. In each location decision, considerations of 
agglomeration and nodality play a role — that is, the size and closeness (essenti- 
ally in economic terms) of the market and of the input sources are weighed. 10 
As a result, the existing distribution of population and economic activities 
among and within the regions in itself becomes a factor in differential regional 
growth. 

The current agglomeration or spatial configuration of the country, which is 
essentially the end-result of past growth, is itself a significant influence on future 
growth. This arises from the fact, discussed earlier, that the existing agglomera- 
tion pattern has a great deal to do with decisions currently made by producers 
as to the appropriate location of new or added production facilities. 

Whenever there is development of a given resource or group of resources 
within a given region in response to external and internal demand, fixed capital 
is invested; and the. rate at which it can be depreciated profitably often makes 
for its continued use in this region even if opportunities for somewhat higher 
rates of return to new capital investment develop elsewhere. Once an investment 
is made in plant and equipment, a firm is likely to find that it pays to continue 
operations within the existing location for a long period of time. 

Location decisions well into the future will be influenced by the fact that a 
number of economic activities have located close to each other at a given period 
of time because of the circumstances of that time — as, for example, when a large 
group of iron and steel and linked metal fabricating plants, together with related 
servicing industries, locate in a given area because of the existing natural 
resource, transportation, and market situation and the general pattern of prices. 
The influence arises from many factors but especially from those which have 
come to be analyzed in terms of agglomeration economies (and diseconomies), 
which are appropriately viewed as external economies arising from the concen- 

10 It should be noted that location decisions are often made in a descending geographic 
scale. A firm producing for the national market will normally decide first on the broad 
region which best satisfies its input and output requirements. Often this will mean a deci- 
sion to locate close to the greatest concentrations of population and industry. At the 
second level (what is sometimes called "localization") it will decide which specific locality 
within this region will best serve its requirements. Thus, a firm which wants to supply 
the vast market of the Manufacturing Belt may decide to locate at some midpoint, perhaps 
in a city in Ohio, or at the periphery, perhaps in a city in Minnesota or Kentucky or 
Virginia or New Hampshire. Finally, it will have to decide on a specific site within the 
locality of its choice (sometimes called "siting.") 

The number of location decisions needed may be fewer, however. For example, if a 
firm is geared specifically to supplying a more limited regional market, rather than a 
national market, it may only have to make localization and siting decisions. A local trade 
or service enterprise (such as a retail grocery) will normally have to decide only on the 
site. 



82/^4 Framework for Analysis 

tration of economic activities in a given area. These, as pointed out earlier, are 
related to the existence within established centers of linked industries, social 
overhead facilities — particularly transport facilities 11 (and more favorable 
freight rates), specialized business services, and the like, as well as concentrated 
pools of labor (particularly skilled labor). 

To the extent that regional and community agglomerations can generate 
immobile economies (as against economies which are available to all firms 
within an industry irrespective of their location), these economies can be a 
source of interregional cost differences that will have the same type of binding 
effect upon location as the existence of immobile resources. Where they are 
recurrent in space, the choice of one of a set of possible agglomerations will 
depend upon the relative interregional price difference they generate and their 
position with reference to other mobile inputs and markets. Since these agglom- 
eration economies tend to coincide with high population concentrations (hence 
markets), their effect often is to strengthen existing tendencies to market orienta- 
tion — and the piling up of more agglomeration upon existing agglomeration. 
The tremendous growth of our metropolitan centers can be explained in large 
part in these terms. 

The distribution of people and economic activities which has evolved from 
the past influences all current location decisions because the overwhelming 
majority of location decisions must take market, input, and transport factors 
as given. Thus, the decision of the present, or the marginal decision in other 
words, is based to an important extent on the locational (and price) situation 
as it has evolved from the past. This is not to suggest that a firm, in making 
its location or production decision, does not look ahead; rather, the main 
element is that, at any point in time, the future costs and returns are necessarily 
estimated in terms of the relatively slowly evolving pattern of agglomeration. 

The same general considerations apply in the investment made in a given 
location by the labor which comes in to help exploit the resource or group of 
resources. Here the investment may be, in part, in the form of fixed capital 
in a home and a different type of investment arising from close bonds to friends, 
children's schools, and the comfort of being among familiar things. There is 
some analogy here with the situation of fixed capital investment. The older 
person, like the older firm, will weigh differential opportunities differently than 
will the younger person with far less investment in the existing location. 

Given the considerations behind locational decisions, it is not surprising 
that regional agglomeration patterns have considerable permanence. For 
example, in 1900 the Manufacturing Belt contained 49% of the population 
and 74% of the nation's manufacturing workers; in 1957 it still had 46% of 
the population and 64% of the manufacturing workers, despite the tremendous 
growth of the Southeast and the West over these years. 

"Clearly, the network established by the railroads in the middle of the 19th Century in 
response to the regional pattern of production and trade of that period gave significant cost 
advantages to firms locating along the main lines, where other production considerations 
were equal. 



Location Factors / 83 

But while the relatively inflexible elements which reflect "sunk" costs must 
be given adequate consideration, there are also significant elements of flexibility 
in the economic system. Capital (and particularly "new capital") as well as 
labor (and, again, particularly "new labor") do tend to seek out new and 
better opportunities and will move over wide areas of the country to achieve 
an improved over-all situation. Free trade among the regions can be expected 
to work strongly towards a geographic equalization of the prices of the factors 
of production — or, at least, towards a narrowing of a spread in prices — through 
its effect on the relative demands for the different factors. 12 However, the high 
degree of flexibility in the economic system is implicit in the entire discussion 
of input-output access and need not be elaborated here. 



Nodality 

The role of nodality, or orientation toward a central place or node, should 
also be touched upon here. Nodality has at least two important features of 
significance for regional economic growth; first, in relation to agglomeration 
and, second, in relation to transportation and transport cost. 

Nodality helps to provide deeper understanding of the agglomeration con- 
cept. In the broadest sense, it suggests the locational relationship of economic 
activities to each other and to population with a focus on the relative degree of 
concentration and distance. 

The importance of nodality for regional economic growth stems from the 
hypothesis, for which a number of scholars have presented strong evidence, 
that closeness to population and industry concentrations directly influences the 
volume, type, and intensity of economic activities. 13 Nodality in this sense can 
be viewed from three standpoints: first, the position of a given area with re- 
spect to the over-all configuration of population nodes or to the national market 
(thus, for example, locations in the western end of the Manufacturing Belt 
have become increasingly advantageous with the growth of the West) ; second, 
distance from a major population node or metropolitan center (the very use 
of the term "center" suggests the nodality element and an expected decrease in 
population and activity as distance from the node increases) ; and, third, dis- 

12 See Paul Samuelson, "International Trade and the Equalization of Factor Prices," 
Economic Journal, Vol. 58 (June 1948) , pp. 163-84 and Svend Laursen, "Production 
Functions and the Theory of International Trade," American Economic Review, Vol. 42 
(Sept. 1952), pp. 540-57. 

"See, for example, T. W. Schultz, The Economic Organization of Agriculture (New York: 
McGraw-Hill, 1953), p. 147. Schultz has suggested that "The existing economic organiza- 
tion works best at or near the center of a particular matrix of economic development." 
Also, Alexander Melamid, "Some Applications of Thiinen's Model in Regional Analysis of 
Economic Growth," Papers and Proceedings of the Regional Science Association, Vol 1 
(1955), pp. L1-L5, and John R. P. Friedmann, The Spatial Structure of Economic Develop- 
ment in the Tennessee Valley (University of Chicago, Program of Education and Research 
in Planning, Research Paper No. 1, March 1955). 



84 / A Framework for Analysis 

tance from a number of minor nodes or the extent of urbanization within a 
given broad area. 14 

Through the use of "nodal indexes" representing the three aspects of nodality, 
the role of space in the development of the economy can be analyzed in terms 
of various significant measures of activity within the various economic sectors, 
such as agriculture, mining, manufacturing, etc. Thus, the intensity of farm 
and mining activities, for example, and the relative returns to these activities 
can be measured by way of nodal indexes. To take just one example, measure- 
ment of this type reveals that access to markets is conducive to intensive 
agricultural activity; but it is access to a regional market, or metropolitan 
center, 15 rather than access to the national market (as measured by population 
potential) that is chiefly associated with high per-acre values of products sold 
and of high per-acre values of land and buildings. 16 In general, the use of the 
nodal measures greatly enriches the concept of agglomeration and provides 

"The first aspect can be examined by way of the measure of "population potential" 
(Stewart) and of "market potential," (Harris) "... an abstract index of the intensity of 
possible contact with markets." Population potential can be regarded as a summary 
measure of the distribution of population over the entire nation with reference to a given 
point in the nation. The value of the potential index reflects not only the density or con- 
centration of population in the local area, but the settlement pattern over the entire United 
States. Theoretically, the potential of population at a given point is obtained as the sum, 
over the entire population, of the reciprocal of each person's distance from that point. In 
practice, of course, the computation is approximate. 

The initial concept was developed in relation to population by John Stewart. See 
"Empirical Mathematical Rules Concerning the Distribution and Equilibrium of Popula- 
tion," Geographical Review, Vol. 37 (July 1947), pp. 461-85; and "Demographic Gravita- 
tion: Evidence and Applications," Sociemetry, Vol. 11 (February-May, 1948), pp. 31-58. 
More recent models based upon these concepts include Chauncy D. Harris, "The Market 
as a Factor in the Localization of Industry in the United States," Annals of the Associa- 
tion of American Geographers, Vol. 44 (December 1954), pp. 315-48; Edgar S. Dunn, Jr. 
"The Market Potential Concept and the Analysis of Location," Papers and Proceedings of 
the Regional Science Association, Vol. 2 (1956), pp. 183-94; William Warntz, The 
Geography of Price, unpublished doctoral dissertation, University of Pennsylvania, 1955. 

The second facet can be indexed approximately by the linear distance separating an 
area from its nearest metropolitan center. The degree of proximity to a major population 
node, of course, influences the area's population potential; but the measures are not 
redundant. An area of relatively low population potential can be located in the vicinity 
of a metropolitan center, for that center is only one of many agglomerations of population 
which jointly determine the area's potential. 

An "index of urbanization," a summary measure of the size-of-community distribution 
of the population residing in the local area, can be used as an indicator of the third facet 
of position in the spatial framework of the economy. The size-of-place distribution of 
population in the local area is formally and empirically independent of the area's distance 
from a metropolitan center and would seem to be a negligible component of the area's 
population potential. 

^This is, of course, in keeping with what would be expected on the basis of economic 
theory; namely, that given the greater competition for land among alternative uses in areas 
close to metropolitan centers, agricultural uses can compete successfully only if they use 
the land intensively and yield a substantial net return to the land. 

16 The relationship between distance from a metropolis and value of products sold is such 
that if, in 1950, a State Economic Area was 130 miles from the nearest metropolis, it could 



Location Factors / 85 

valuable insights into the role of space in regional economic growth by pointing 
up specific locational relationships. 

The critical element in location, as has already been suggested, is the rela- 
tive accessibility of the various parts of the country to inputs and markets. 
Accessibility, however, is less a matter of distance in miles than of relative 
transportation costs. This phase of the problem has already been touched upon; 
we need only note here that it tends to reinforce the evolutionary growth features 
of the economy. 

Historically, it was no accident that cities tended to develop in places where 
transportation by water afforded a relatively cheap means of exchanging com- 
modities. Later, as means of land transportation were developed, these tended 
to connect areas which already had relatively high concentrations of production 
and trade. Today, because these areas offer a wide variety of modes of trans- 
portation, with favorable freight rates based on volume and special arrange- 
ments, they are often favorable sites for the assembly of raw materials or 
partly fabricated inputs and for the distribution of intermediate or final outputs. 
It is useful, in this connection, to re-examine the "growth" areas (as discussed 
in Chapter 2) in terms of their location with regard to the established major 
transportation routes. Given the role of input-output access, it is of course 
no surprise to find that areas advantageously located at breaks-in-transport 
points reveal important advantages for economic growth. 17 



LOCATION OF CONSUMPTION ACTIVITIES 

Production activities seeking locations have a two-fold interest in populations 
and their regional distribution : as a source of labor inputs and as a destination 
for outputs. Presumably, then, any factor influencing population distribution 
would also influence the pattern of production activities. 

In analyzing regional population distribution, it is convenient to view the 
household as the decision-making unit that assembles consumption inputs (con- 
sumption products and services) and distributes labor outputs. When the house- 
hold is viewed in this way, it appears in the first instance that the transfer costs 
in marketing its labor output over any considerable distance exceed those of 
assembling its consumption inputs. In short, the household is oriented to 
regional markets for its labor output. It would follow that populations would be 
distributed essentially on the basis of the regional distribution of economic 
opportunity. Therefore, the factors that explain the location of production 
activity would explain the distribution of consumption activity. 

be expected to have an average value of product sold per acre of about $15. On the other 
hand, an area which was 35 miles from the nearest metropolis could be expected to have 
an average value of product sold per acre of about $32. Of course, there are many SEA's 
showing marked deviations from the expected value based on the formal measures. 

^The "nodality" facet of the study has been investigated by Duncan and his associates 
and the results of their research appear in the volume Metropolis and Region. 



86/^4 Framework for Analysis 

This does not cover the total picture, however. An important part of our 
population is not engaged at all in the marketing of labor outputs. Some 8% 
of the U. S. population is over 65 years of age. Approximately two-thirds of 
these persons are not working, and many enjoy some form of paid retirement. 
They are free to seek locations that give them greatest access to the consumption 
inputs they feel are important. Since many of their basic consumption require- 
ments can be met equally well in all regions, with only minor differences, many 
of these persons will seek out the more intangible advantages (such as climate 
and coast) that do vary substantially from region to region. 

In addition, many households that do provide labor outputs may find that 
good transfer relations with reference to consumption inputs are more important 
than good transfer relations with reference to labor markets. There are families 
that will choose to move to large urban centers in order to take advantage of bet- 
ter educational and health and welfare facilities, even in the face of uncertainty 
with regard to job opportunities. The force of consumption considerations is 
apt to be particularly strong over shorter distances. The movement to the 
suburbs can be interpreted in part at least in terms of the search of households 
for more convenient access to certain consumption inputs at the penalty of 
significantly greater transfer costs in marketing labor outputs. Specialized 
services often can be supplied over substantial distances, allowing the household 
to be located in a place that is considered to give it greatest access to its con- 
sumption requirements. 

Still further, certain firms that are oriented to labor markets tend to be foot 
loose. In the absence of compelling transfer cost factors affecting production 
location, these firms may often seek locations that provide favorable access to 
the kind of consumption requirements that management or some employee 
consensus may deem important. 

It is difficult, given the lack of information on this point, to evaluate the extent 
to which the location of populations is essentially independent of considerations 
affecting the location of production. However, while the location of jobs seems 
clearly to be the critical element, some attention must be given to the role of 
"independent" household decisions in analyzing regional economic growth, 
particularly for areas like Florida, California, and parts of the Southwest and 
Mountain regions. It is not surprising to find that the most rapid economic 
growth in recent decades has taken place in precisely those areas where both the 
consumption and production elements are relatively favorable and reinforce 
each other. 



J / ^Input-output access 
and multipliers 



;? 



"Input-output access" is a netting-out of the relative advantages and dis- 
advantages of each region for the economic activities of any given industry 
or of all industries combined. Because the input and output requirements vary 
greatly as between industries (as is evident when the requirements for agricul- 
ture and for manufacturing are compared), the netting-out must necessarily be 
selective and related to specific industries rather than in terms of a measurable 
aggregate quantity or single index. 1 

When we speak of access as the sum of the relative advantages and dis- 
advantages for the production of a particular commodity at some given place, 
we refer to more than just the resistance, and hence costs, imposed by space on 
the assembly of inputs and the distribution of outputs. We refer, rather, to all 
the cost elements inherent in production of the commodity. Transport costs, 
to be sure, are a significant proportion of total costs in many cases, but they 
are never the only element in cost. The prices of inputs and final products for 

lr There are a number of reasons why a general index of access cannot at this time be 
utilized in explaining regional growth and anticipating the growth potential of regions. 

(1) Access as an element in economic growth is basically functional and specific. For 
example, a region might have a low access index relative to the aggregate of basic resources 
but a very high access to one or two specific resources (say, oil and gas) capable of support- 
ing much activity. In the same way it might have a high access to specific markets. 

(2) In a modern economy such as ours, the processing stages have become so subdivided 
that a region may not need very high access to natural resources and final markets to sup- 
port significant growth. The important thing may be access to specific intermediate sup- 
pliers and markets. 

(3) Regions would seem to differ in the degree to which their access potentials are 
realized. A region with a relatively low over-all access potential that has been only partially 
realized can be expected to show more growth over a given time period than another region 
with a higher initial potential already fully exploited. 

(4) In a modern economy — where many decisions are consumption-oriented — free re- 
sources such as climate are significant to economic growth, and there is no practical way to 
include these in a general measure of access. 

The main point we wish to stress is that, to come to grips with the problem of access, it 
is necessary to move to a consideration of specific functions in specific regions. 

/ 87 



88 / A Framework for Analysis 

a particular firm or industry at a particular point in space are also influenced 
by the competition of rival producers or products. Likewise, a "good" oppor- 
tunity at a given place might not be exploited because of the existence of a 
better opportunity elsewhere. Therefore, "rivalry" and "opportunity" costs, 
as well as transport costs, are included in the concept of access. 



Returns on New Investment 

So far, we have referred only to the relation between price per unit of output 
and the prices of factors of production, both material and labor, and to techno- 
logical relationships at the point of production. But when the access concept is 
used as a framework for studying economic growth, another distinction must be 
made. One of the critical elements for economic growth in an area is new 
investment, and decisions on new investment are determined by relations 
at the margin — that is, by small increments of change rather than by average 
relationships. Thus, it is quite possible that an area might have, on the average, 
favorable conditions for the production of a given commodity and not grow 
simply because the opportunities for new investment are unfavorable in this 
area relative to other areas. 

To illustrate this, let us consider three different regions and summarize the 
factors affecting the level of investment in a particular industry in these areas 
by three schedules, one for each region. This we do in Figure 20, on which we 
plot for each region the marginal return per dollar of capital stock employed in 
production, r, as a function of the total capital stock, K. Suppose further, for 
the purposes of this illustration, that there is a level of return, r e , at which new 
investment is just profitable and that this rate is the same for all three regions 
and independent of total capital stock in the region. The total capital stock 
that each of the three regions will support under the conditions summarized in 
the schedules of marginal return on capital are K 1? K 2 , and K 3 , respectively. In 
equilibrium, defined as a position in which there is no incentive for anyone 
to make new investments, the total capital stock in the first region would greatly 
exceed that in either of the other two. Suppose, however, that, at a given point 
of time, capital stock in Region 1 is K x while in Regions 2 and 3 it is K 4 . Such 
a situation would imply that the rate of return on new investment in the latter 
two regions would be r l5 which exceeds r e , and hence it would be profitable to 
add to the capital stock. In Region 1, however, there would be no incentive 
for anyone to add further to the total capital stock. Hence, the growth possibili- 
ties are better for Regions 2 and 3 than for Region 1. 

But Figure 20 also illustrates another point. Given that the rate of return 
on new capital investment in Regions 2 and 3 is the same, it does not neces- 
sarily follow that the growth prospects for the two are the same. As the figure 
indicates, because the schedule for Region 3 is less steep than the one for 
Region 2, it will take a greater increase in the capital stock in Region 3 to 
bring the marginal return on new investment down to the point where no 
further investment is profitable. The slopes of these schedules depend, among 



"Input-Output Access" and Multipliers / 89 

other things, upon the extent to which price received per unit of output falls and 
price paid per unit of material and labor input rises with an increase in the 
region's output. (Changes in transport costs which might accompany changes 
in the volume of material input shipped into the region or of output shipped 
from the region influence the extent to which input and output prices change 
as the region's output expands.) 

One further point. If we are seeking to explain differences in growth over 
some time period, such as 1939 to 1954, we might interpret Figure 20 as show- 




H"2 



Figure 20. Returns on Capital in Three Regions. 



ing the schedules of marginal return on capital, given the conditions that exist 
at the beginning of the period. As discussed above, given these conditions, 
prospects for growth at the beginning of the period depend upon the actual 
capital stock in relation to the equilibrium capital stock. However, over a 
rather long time period such as this, these schedules would shift as a result of 
shifts in demand for the region's output, changes in technology, and shifts in 
the supply schedules for the inputs used in production in the region. (Shifts in 
transport costs might lead to shifts in either the demand for output or the supply 
of one or more material inputs.) Thus, in addition to the factors discussed 
above, shifts in the schedules in Figure 20 would lead to differential growth 
rates as among regions. 



Consequence of Variations in Access Characteristics 



When regions are examined in terms of input-output access (as defined 
above) with regard to the requirements of specific industries and for all 
economic activities taken together, the extent to which they vary in their 



90/^4 Framework for Analysis 

prospects for growth becomes apparent. Conceptually, and in an abstract 
sense, one might use the variations in access characteristics to identify at least 
sixteen possible types of regions, as presented in Figure 21. Clearly, the table 
is an extreme simplification and has only general illustrative meaning. Also, 
"good" and "poor" are only relative designations and obviously do not repre- 
sent discrete and separate types. But a classification scheme such as this, though 
artificial, has the advantage of focusing attention on the wide range of prospects 
for growth. For example, one cannot think seriously about the basic concept 
and not see the fallacy of the extreme local economic-development approach 
which regards every region and community as capable of limitless economic 
expansion. 

To the extent that a region's general access characteristics may be taken as a 
rough index of its potential for growth, Region 4 in Figure 21 would have 
little prospect for growth, while Region 13 would have an unsurpassed growth 
potential. 

Other regions fall in between these extremes. Regions 1, 2, 3, 8, 12, and 16 
are only a little better off than 4. In these cases, reasonable access to inputs is 
offset by lack of market access, or vice versa. Thus, for example, one type of 
region may be developed to the extent that it contains an important mineral 
resource which is much in demand, but its development may be essentially 
limited to the exploitation of that particular mineral because of its disadvantage 
with regard to transporting almost all other products to distant markets. In 
general, regions 6, 7, 10 and 11 are somewhat better off because they at least 
have some access to both inputs and markets. Regions 5, 9, 14, and 15 are 
still better off because they combine good internal access to markets or inputs 
with good external access. 

The restraints placed upon the future prospects for these regions also differ. 
Region 4 would have very dim prospects of evolving into a region type with 
greater growth potential. Any change in its character must rest upon a doubly 
fortuitous set of circumstances. Technology, discovery, or institutional changes 
must bring about an improvement in its access to both inputs and markets. 
For Region 3, which has good access to inputs in the home region but no 
external markets, the prognosis is poor but not as hopeless. This region must 
either (a) overcome the restriction upon its transfer relationships with other 
regions or (b) exploit its resources through a prolonged series of internal 
growth sequences. Region 12, with good access to external markets but poor 
access to input sources, might have a better prospect for breaking out of its 
dilemma. Discovery, technology, or even the pressure of growing demand might 
improve its access to basic inputs. 

A major advantage of this schematic approach is that it forces attention on 
the variety of growth experiences that can be found in regions of different 
types. This is especially true if we consider the direction of growth sequences. 
As suggested earlier, it is widely assumed that development tends to follow a 
prescribed sequence, with growth initiated by advances in primary extractive 
activity. However, this is not always the case for the regions of an economically 
advanced nation. 



'Input-Output Access" and Multipliers / 91 









Good access to 


basic inputs* 


Poor access to basic inputs* 








from external 


regional and 


from external 


regional and 








national 


sources 


nationa] 


sources 








Good access to 


Poor access to 


Good access to 


Poor access to 








basic inputs in 


basic inputs in 


basic inputs in 


basic inputs in 








home region 


home region 


home region 


home region 


-d 


O V 












c 

eo 


o 

CO JC 














co 


CI 

o 

'ti 












^ .5 

o 
eO 


#1 


#2 


#3 


#4 


'3b 


10 en 


0) 










0} 


o H 












■-J ^ 


O d 












CO fcH 

a «3 


Oh £ 




II 


I 


I 





S3 S 




























x •— i 














flj Co 














fl 


o a; 












o .2 


~ S 












cO 


o 












a) R 


!/--■ — 

CO 
















CI 

.2 


#5 


#6 


#7 


#8 


o 


rt « 


'5c 










CO 


co 

"5 


u 










M 


T3 M 












o 


O fa 












o 


O rt 












Dh 


o s 




III 


II 


II 


I 


n3 


o <u 












c 


~ 6 












CO 


o 

to -^ 














(O 


c 
.2 
'Si 
a; 
u 










CO 

a 

.2 

"Sb 




#9 


#10 


#11 


#12 


co 


O >i 












~Z ^ 


o 5 












CO h 

CI CO 


Ph S 




III 


II 


II 


I 


<D C 




























X .— 1 














<L> CO 














fl 


O V 












o .2 


*• s 
























CO 


o 












CO C 


CO 












CO 
CD 

o 




CI 

o 


#13 


#14 


#15 


#16 


O 
CO 




"5c 

u 










-T3 


T3 M 












o 


O fa 












o 


O CO 












o 


o s 




IV 


III 


III 


II 



*Not only basic resources but important intermediate sources need to be considered. 

Note: Roman numerals indicate number of "good" access dimensions, and suggest relative 
over-all locational advantages or disadvantages. 

Figure 21. A Schematic Presentation of Types of Regions that Can 
Exhibit Different Potentials with Respect to Growth. 



A region such as type 11 might show a sequence of development completely 
the reverse of the sequence typically hypothesized. Florida is an example. It 



92/^4 Framework for Analysis 

has had relatively poor access to home markets (that is, a limited scope for 
development on the basis of the size of the home market) and to external input 
sources. The major characteristic of its input access in the home region has 
been its large fund of coast and climate. Its access to external markets for this 
resource was good because, in our highly developed economy, population 
movements often take the form of a quest for amenities rather than economic 
opportunity. The exploitation of a resource was dominant in this development, 
but it is a special kind of resource that might be identified as a resource-service. 
No primary activities in the old sense were associated with its exploitation. 
Rather the exploitation of this resource required an intense development of 
tertiary activities servicing population. In 1950, some 66% of all employment 
in Florida was in tertiary activities. Market-oriented activities dominated. 

A stage is developing, however, where the population concentration and the 
availability of business services are attracting increasing quantities of secondary 
manufacturing activities. Typical of these are small-scale market-oriented manu- 
facturing such as metal construction products and relatively footloose activities 
such as electronics. As the wealth and size of the population grows, deficit food 
supplies make possible the use of agricultural lands that, at an earlier date, 
could not be considered an economic resource. This calls for an expansion in 
the primary sector of the economy. In this situation the tertiary-secondary- 
primary sequence is more logical than the reverse. 

Actually it is not at all impossible for a growth sequence to start in what is 
thought of as typically the "middle" and perhaps go both ways. A region of 
type 6 or 8 might exhibit this kind of sequence. Sometimes the exhaustion of 
a resource or the development of a substitute may leave a region "over- 
developed." If labor is slow in out-migrating, the pool of relatively immobile 
labor with depressed wages may attract secondary manufacturing activities 
oriented to cheap labor. The new secondary activity may induce growth se- 
quences that lead to expansion in primary, tertiary, and other secondary activi- 
ties in the region. If this should happen, the region may regain a level of 
growth and production commensurate with that of other regions. In short, in 
an advanced society such as ours, there are many different kinds of regional 
growth sequences. Growth may begin with primary, secondary, or tertiary 
activities and it may proceed in several directions. 

The variety of growth experience is apparent not only in the different 
sequences that are possible but also in the variety of functional pathways it 
might take. For a region (such as 7, for example) with good access to inputs 
and markets only in the home region, growth is largely restricted to the internal 
evolution of specialization characteristic of more or less "closed" regions. In 
another region (such as 11) growth may take the form of interregional special- 
ization in response to external stimuli. In other regions (such as 13) it is more 
apt to be compounded of elements of both external and internal response. 

A region (such as 10) which has poor access to inputs and markets in the 
home region might sustain considerable growth because of its nodal position 
with reference to external sources and markets. 

Regions vary widely in their capacity to achieve "mature" development. A 



"Input-Output Access" and Multipliers / 93 

rationalized, variegated, mature development of economic functions is unlikely 
in a region that does not have good access to external national markets. All of 
the functions that are dominated by important external and internal scale 
economies would be denied to it. 

In general, then, interregional differences in access to inputs and markets can 
be expected to bring about different patterns of regional growth behavior. Con- 
versely, the various growth patterns displayed can be explained by identifying 
the relative advantages and disadvantages of the regions with regard to input 
and output access for the major types of economic activities. 



Multipliers 

Regional economic growth (or decline) is not simply a consequence of a 
discrete set of locational decisions. Every sizable injection of new investment 
within an area or increase in inputs and outputs, brings in its train a series of 
related economic effects. The economic impact of such forces upon an area is 
encompassed within the concept of regional multipliers. A highly complex set 
of interrelationships are involved. Thus, various autonomous forces can be 
considered to come into play: new investment; changes in the volume and 
composition of inputs (particularly labor) ; changes in the volume and composi- 
tion of output of the area's industries; and, cutting across the latter two, changes 
in the volume and composition of exports and imports. The economic impact 
can be viewed in terms of the effect on employment in the area, on income, on 
household purchases, and also on induced investment. 

The dynamics of the multiplier is one of the reasons why regional growth is 
not simply a matter of the proportion of "growth" industries within a given 
area or of the ratio of export to residentiary activities. While specific measure- 
ment is difficult, 2 the evidence suggests that there is no direct correlation be- 
tween the rate of growth of a given industry and its multiplier effects in terms 

2 Although highly suggestive approaches to the measurement of regional multipliers have 
been proposed, the concept is not sufficiently developed nor the data adequate to permit 
us to work out actual impact measures. Frederick T. Moore and James W. Petersen, 
employing a modified Leontief interindustry relations model, have attempted to estimate 
the multiplier effects within a state. "Regional Analysis: An Interindustry Model of Utah," 
Review of Economics and Statistics, Vol. 37 (November 1955), pp. 368-83. 

Werner Hirsch has developed an interindustry flow table for metropolitan St. Louis, 
employing 1955 data from company records, on the basis of which he has worked out 
estimates of income and employment multipliers. "Interindustry Relations of a Metropoli- 
tan Area", Review of Economics and Statistics, Vol. 41, August 1959. Thus, he finds, for 
example, that the income impact of the printing and publishing industry on the St. Louis 
economy in 1955 was such that a $1 million final demand increase raised the area's income 
by about $870,000. By contrast, the same amount of increase in final demand in the 
products-of-coal-and-petroleum industry could be expected to bring about an increase in the 
area's income by only some $220,000. The extremes in the case of the employment 
multiplier, Hirsch finds to be Plumbing and Heating Supply and Fabricating Structural 
Metal Products (an employment gain of 138 workers per $1 million final demand increase) 
and the Products of Coal and Petroleum (a gain of only 14 workers). 



94 / A Framework for Analysis 

of inducing additional economic growth within the area in which it is estab- 
lished. Also, export industries do not all generate equal amounts of residentiary 
or service activity. The various types of multipliers are related to different 
kinds of industry characteristics. It is clear, for example, that a given industry's 
employment multiplier will be directly related to its own relative labor-intensive 
character and to the character of its "backward-" and "forward-linked" in- 
dustries attracted into the region, and of course labor intensity has a logic of its 
own, not necessarily correlated with "growth" measures. 

As a shortcut in dealing with this highly complicated set of phenomena, it is 
convenient to think of the growth or decline set in motion by any activity 
change in a region as depending upon what might be called its activity inter- 
action potential. This refers to the nature of the input-output ties which an 
activity has with other activities within the same region and elsewhere. Given 
an activity change, the kind of growth (or decline) that will follow, and its 
duration, depend upon the extent to which the new activity is closely tied to 
(or independent of) the associated processes that precede and follow it. For 
example, a hosiery mill attracted to a given area by surplus labor and a mini- 
mum of community services within this area will, by itself, normally generate 
little subsequent growth within the area. Its supply sources and markets are 
such that its input-output ties are almost entirely external. In conjunction with 
other activities, it might make a modest contribution to regional population 
growth and add modestly to the demand for community and business services. 
By contrast, a new steel mill, such as the Fairless plant on the Delaware, may 
bring with it large population migrations, huge construction expenditures, many 
associated fabricators and suppliers, and generate multiple rounds of expansions 
in community and business services and marketing functions. 

Multiplier effects are related to the technology and organization of a given 
industry at a given period of time. The time element is important because tech- 
nology and organization influencing activity interaction within any given in- 
dustry is subject to change, and in some industries — such as the chemical 
industry — rather rapid change. 

Multiplier effects also have an over-all component, aside from the individual 
industries. This is related in an important way to how highly developed or 
relatively underdeveloped a region happens to be. In a less developed area — 
where social overhead facilities, such as public utilities, have to be built up and 
new services have to be brought into being — a given amount of manufacturing 
growth will generate more economic expansion than it will in a "mature" 
region which already is well equipped with facilities and services. In the 
latter case, there is likely to be a certain amount of unused capacity that can be 
employed to service the needs generated by the manufacturing growth. 

As population grows regionally and nationally, some regions will reach a 
threshold where certain activities become feasible for the first time. Every 
activity has some minimum-sized market that its scale economies require for 
efficient production or distribution. Every region has a degree of market con- 
centration and nodality that establishes limits upon the kinds of internal and 
external economies that are available. In a region where the development of 



"Input-Output Access' 9 and Multipliers / 95 

economic activities is relatively undifferentiated, population growth may push 
that region to successive points where new functions become feasible. 

It also follows that these new functions serve as a special stimulant to growth. 
Regions introducing them would claim a greater advantage in growth relative 
to (a) regions that had not yet achieved the threshold and (b) regions that 
would not expect to expand their functions much further. The areas most likely 
to achieve these successive thresholds are those that (a) have already achieved 
some minimum degree of accumulation of population and (b) are already 
serving as foci for a pyramid of regional and metropolitan functions. 

This set of factors would seem to be particularly important in explaining the 
dramatic growth of California and, to a lesser degree, of the other states of the 
Far West and the Southwest. The multiplier effects here are reinforced by 
relative isolation from other important national centers of activity. Under such 
circumstances, activities that possess only marginal efficiency cannot be as 
easily provided by adjacent regions during early growth periods. This helps 
to explain why certain of the economically undifferentiated sectors of the country 
seem to be outrunning the older sectors in relative (and sometimes absolute) 
growth rates. Henry Bruton puts it this way: "Increasing size of an economy 
proceeding from a small base results in various kinds of economies of scale, 
but such economies are finite in quantity and effect. It seems clear that, after 
an economy has reached a given size, further increases in the size of the system 
itself will not result in further economies of scale and indeed may create a 
classical diminishing returns problem." 3 

Multiplier effects are intimately tied to the sequences of economic growth 
within a region. (This proposition can equally well be put the other way 
around.) This can be highlighted by an illustration. 

Hypothesize the initiation of exploitation of an important new resource 
deposit in a relatively unpopulated region. (It would not matter whether 
technology or discovery created this new resource, or whether it was due to a 
significant change in relative prices.) This might well lead to a series of cumu- 
lative changes, somewhat as follows: 

(1) Initial investment in resource facilities accompanied by an initial in- 
migration of workers and their families would take place. (2) Improvements 
in internal transport facilities and new transport links with other regions would 
follow. (3) Community facilities and services would be provided, such as 
housing, water, sewerage, and so on. (4) The new population nucleation would 
attract some small market-oriented enterprises. (5) The initial resource ex- 
ploitation would attract suppliers of materials and equipment as well as busi- 
ness services ("backward linkage"). (6) The resource source might also at- 
tract some input-oriented processing enterprises. (7) The requirements for 
new construction and new employment in basic production and related umbili- 
cal activities, in community facilities, and in population-oriented activities 
would involve further expansion in population and new enterprises supplying 

3 Henry J. Bruton, A Survey of Recent Contributions to the Theory of Economic Growth 
(Cambridge: Center for International Studies, Massachusetts Institute of Technology, April 
1956) , p. 148. 



96/^4 Framework for Analysis 

materials and services for the construction activity, so that the initial agglomera- 
tion of people and activities grows in size and complexity. (8) It now becomes 
possible for some activities to move in that are relatively footloose but do 
require some minimum of community and business services to function effi- 
ciently. (9) There is further growth in general and specialized business services. 
(10) As the population and industry complex continues to grow, the transport 
relations with other regions are improved and refined, and this elaboration of 
the inter-regional network may lead to the new community developing some 
new "nodal" functions in transfer relationships with other regions. It is possible 
that the change in transport configuration might attract to it some enterprises 
that are "intermediate oriented." (11) The growing size of the population 
complex can attract more and more activities whose scale economies require 
larger market areas. (12) As the region grows, specialized services, market 
institutions, and financial sources evolve, and the region becomes more and 
more attractive for enterprises requiring these services. (13) The basic changes 
in the location of inputs and markets brought about by the development of the 
region will alter production and consumption relationships in other regions and 
thus transmit to them elements of its dynamism. (14) Increases in external 
demand will stimulate production within the industries established in the region. 
(15) And so on. 

Note, however, that though this sequence is logical, it is not necessary. Its 
form might have been quite different if the new resource had more limited ties 
with market-oriented input sources and input-oriented markets, if the popula- 
tion movement in response to the new economic opportunity had been blunted 
by a regional oversupply of labor and only persons already in the region had 
been absorbed in the resource activity, if the combined multiplier effects did 
not bring the region to the threshold where scale economies would make 
feasible a substantial diversification of activities but stopped at the directly 
linked industries, etc. And of course the whole picture would have been very 
different if the resource exploitation had been undertaken in a region already 
highly developed and possessing an extensive complement of services and 
facilities. 

This brings us, round circle, to the initial proposition that regional economic 
growth can be viewed most effectively in terms of (1) the more important 
stimuli, particularly as evolving from structural changes within industries and 
among industry groups in response to major market and technological changes, 
(2) the characteristics and functional composition of the various regions as 
potential "carriers" of the changes, and particularly the evolving "quality" of 
access to inputs and markets and the multiplier potential, and (3) the basic 
types of growth sequences which are generated. This involves a recognition 
of the rather wide variations in regional economic growth patterns that can be 
expected to result from the highly complex characteristics of industrial change 
and from the basic differences in regional structure. 




Factors influencing per capita incomes 



The index that has come to be generally accepted as the best over-all measure 
of national economic growth is a sustained increase in per capita income. There 
is a clear logic in this. In the first place, any measure of growth among nations 
of different sizes, say China and Cuba, can more readily be used when population 
size is taken into account. In addition, a continuing increase in total population 
within a nation from year to year and over long periods of time can normally 
be expected. Given the severe restrictions on international migration, the key 
"volume" item of population increase can thus be taken for granted. Increase in 
total income, under such circumstances, now has a minimum rate or floor 
attached to it; psychologically it is hard to regard an increase in the volume of 
income which is at a lower rate than the increase in total population as "econo- 
mic growth," although in the volume sense, growth is taking place. Thus, it is 
understandable why, in the case of national economies, increase in per capita 
income should be taken as the measure of economic growth. 

This now traditional approach to the measurement of economic growth tends 
to be employed in the study of regions also. But there are significant differences 
in the case of regions. Continuing population growth cannot be taken as a 
starting point. Families can migrate freely from one region to another, so that 
a given region might well experience a decline in population. In such a case, 
even if total income should remain at the former level, or slightly drop, per 
capita income can still go up, that is, workers' earnings can go up in the face of 
declining population. In other words, a divergence between gain in volume and 
gain in welfare is conceivable — and, as a matter of fact, such divergence has 
taken place in the United States. 

Once it is understood how widely regional growth patterns can vary, it is not 
surprising to find that changes in the industrial sectors — agriculture, manufac- 
turing, etc. — are associated quite differently with volume changes than they are 
with welfare changes. In the previous sections, we have discussed the volume 
forces; here we shall touch upon the welfare forces. 

This can be seen by examining the relation of changes in population and total 
income (the volume measures) as well as in per capita income (the welfare 
measure) with the changing relative importance of different kinds of employ- 
ment by states. In Table 15 we have done this for the two periods 1880-1920 
and 1920-1950. 1 

a The income and population variables used for the comparisons are end-of-period value 
as a percentage of beginning-of-period value. The employment measures are end-of-period 

/ 97 



98 / A Framework for Analysis 

Table 15. Summary of rank correlations: economic growth and changes in 
employment structure, among the states, 1880-1920 and 1920-1950 



Measure 




1880-1920 


1920-1950 


Volume measures: 


Agriculture 






Population 

Total personal income 




+ 0.25* 
+ 0.04 


+ 0.06 
-0.24 


Welfare measure: 




S 


S 


Per capita income 




-0.45** 


-0.48** 


Volume measures: 


Manufacturing 






Population 

Total personal income 




-0.23 



-0.32* 
+ 0.09 


Welfare measure: 




S 


^ 


Per capita income 




+ 0.46** 


+ 0.57** 


Volume measures: 


All Other (Services, 


etc.) 




Population 

Total personal income 




+ 0.08 
+ 0.11 


+ 0.26* 
+ 0.28* 


Welfare measure: 

Per capita income 




+ 0.27* 


+ 0.07 



*Significant at 5% level. **Significant at 1% level. 



During these two periods, there was little relationship between differential 
changes, by states, in population and the volume of economic activities (symbol- 
ized by change in total personal income) on the one hand and changes in the 
employment structure on the other. A close association between changes in 
"volume" within the states and changes in the relative importance of employ- 
ment in the various categories (agriculture, manufacturing, services) would be 
shown by a high rank correlation coefficient. Actually the correlation coefficients 
are neither high nor consistent as between the two periods examined. 2 

But for both periods there were significant and rather large correlations 

employment in the particular category relative to total employment for the state minus the 
same for the beginning of the period. The comparison was made separately for two time 
periods, 1880-1920 and 1920-1950; the latter was based upon forty-eight observations, the 
former on forty-six since income data for 1880 are not available for Oklahoma or separately 
for the Dakotas. For each period the comparison was made for three separate employment 
categories: Agriculture, Forestry and Fisheries; Manufacturing; and All Other (Service, 
Trade, Government, etc.) . 

2 Even in the instances in which the observed rank correlation coefficients are large enough 
to rule out the hypothesis of chance association, they are not large enough to be of much 
practical import. 



Factors Influencing, Per Capita Incomes / 99 

between differential changes in per capita income (welfare) and the changes in 
the relative importance of agricultural and manufacturing employment by states. 
Where the percentage of labor force employed in agriculture tended to decline 
most and the percentage in manufacturing to increase most, per capita income 
increased relatively most. However, no consistent or strong association is seen 
between differential change in per capita income within the states and relative 
importance of the All Other category, which includes services, government, 
trade, etc. 

Thus, sectoral change would seem to be significant in regard to the influence 
of a decline in agriculture and an increase in manufacturing on per capita 
income (for reasons which will be explained later), but to have little to do with 
the differential growth in the volume of regional population and activities. 

On the "volume" side, these data reinforce our view that growth can be 
understood only in terms of the varying input and market requirements of 
specific industries (with the industry disaggregation carried as far as needed to 
differentiate the critical requirements). At the same time, they suggest that 
rather significant insights into welfare can be obtained from an examination of 
even extremely aggregate industrial sector groupings (that is, as aggregate as 
"agriculture" and "manufacturing"). It would seem that the characteristics of 
these broad groupings are more general with regard to returns to labor and 
capital than with regard to their locational and production requirements. We 
find even further, and not unexpectedly, that the industrial structure of a given 
region emerges as being even more determining of per capita income levels when 
structure is disaggregated below the very broad categories employed in Table 
15. 3 Thus, the industrial location shifts, which have been of central concern 
in our discussion of regional growth in volume, can be seen to be highly sig- 
nificant in per worker returns and therefore in regional levels of per capita 
income. In fact, the most critical factors in differential levels of per capita 
income among regions appear to be (a) the relative per worker returns of the 
dominant industries within the various regions on the one side and (b) the 
relationship between capital inflow and population growth on the other. Before 
turning to these, it is useful to note certain characteristics of the welfare 
elements. 



Reasons for Differences in Per Capita Income Levels 

To introduce the discussion of the forces making for per capita income differ- 
ences among different places, let us hypothesize a position of long-run competi- 
tive equilibrium, defined so as to imply among other things that no producer 
would have any incentive to change his rate of output or location and no worker 
would have any incentive to change jobs at a given location or move to a new 
location. Then, let us ask, would we expect to find equality of per capita incomes 
as among regions (however the latter are defined) ? Such a position might 

3 See Harvey S. Perloff, "Interrelations of State Income and Industrial Structure," Review 
of Economics and Statistics, Vol. 39 (May 1957), pp. 162-71. 



100 / A Framework for Analysis 

never be observed, of course, but it might be considered "ideal" in the sense that 
if it were achieved, no transfer of resources could increase the national product. 
Equilibrium in this sense would imply that workers of identical skill would earn 
the same wages everywhere. 

However, even if such were the case, there are many reasons why we might 
expect to find differences in per capita incomes in different parts of the country. 

First, there are several reasons why workers of different groups would receive 
different real wages in such an equilibrium. These are summarized in the obser- 
vation that (1) the relative proportion of skilled and unskilled workers can be 
expected to differ among regions because the employment structures differ, and 
(2) factors other than the current real wage influence the relative attractiveness 
for workers of different employments. Hence, if the industrial and/or occupa- 
tional composition of the labor force differed in different places, one would 
expect these differences to be reflected in different per capita incomes. 

Second, in addition to differences in the price per unit received from the sale 
of labor services, the incomes of persons in different places depend upon their 
ownership of non-human factors of production. Their incomes thus reflect the 
present returns on past capital investments. 

Third, populations at different places differ in the proportion employed. Such 
differences result not only from differences in the proportions of persons not 
normally in the labor force — the young and old — but also from differences in 
the proportions of persons of working age in the labor force. 

Also, there are several reasons why a given level of per capita income would 
have different implications for consumer welfare at different places. If the level 
of money prices of consumer goods were to vary, a given level of money income 
would command different bundles of real commodities in the market. Likewise, 
not all goods and services that affect the welfare of consumers are bought and 
sold on the market. Certain areas have advantages as to climate and other living 
conditions which might imply that a lower than average level of money income 
would attract persons to them. 

Thus, areal differences in per capita money incomes might well persist even if 
the economy were in a position of long-run competitive equilibrium. Such 
income differences need imply neither differences in consumer welfare nor dif- 
ferences in the marginal productivity of labor (as reflected in the level of 
wages) at different places. Likewise, as economic conditions change over time, 
the forces discussed above might lead to continued differentials in per capita 
incomes in different parts of the country. 

In the real world, of course, differences in wages of workers of identical skill 
are actually an important force making for differences in per capita incomes. 
First, the data available suggest that on the average for the nation, and particu- 
larly in the Southeast, the real returns to labor employed in agriculture are 
below those in other employments. Even under static conditions in which the 
total population and the total capital stock were to remain unchanged over 
time, a reshuffling of people from farm to non-farm employments would lead 
to an increase in the national product and to an increase in per capita incomes, 



Factors Influencing Per Capita Incomes / 101 

and hence would lead to economic growth. 4 Likewise, it would appear that 
wages received by workers in a given industry differ regionally and that these 
differences are too great to attribute to equalizing differences (such as differ- 
ances in skills, money prices, etc.). Here, too, a reshuffling of resources would 
lead to economic growth in either the "volume" or "welfare" sense. 

Given the fact that wages are relatively low in some or all industries in a 
particular area, how would we expect this situation to influence the composi- 
tion of economic activity in that area? Economic analysis would suggest two 
things: First, in areas in which wages are below the national average, conditions 
would be especially favorable for location by industries for which labor is a 
relatively important production factor. Hence, we would expect to find some 
concentration of the labor-intensive industries there. And, to the extent that 
the wages of workers of a given skill tend to be below average in an area, firms 
would have an incentive to use more labor relative to capital and perhaps to 
materials in the area. Thus, a low wage area would tend to be one with a low 
capital-labor ratio in all or most industries. 



Influence of Population Growth 

Here we would introduce a final proposition. The regional level of per 
capita income and the differential rate of growth in per capita income funda- 
mentally depend on the evolving relationship between the rate of capital input 
and material input, on the one hand, and the rate of population increase (or 
decrease), on the other. In other words, the implications of the capital-labor 
ratio have to be broadened to include the dynamic factors of capital formation 
and movement among regions and the increase in total population numbers. 

A convergence over time in the levels of wages for workers of equivalent skill 
would tend to bring about a partial convergence of per capita incomes in differ- 
ent places. We say partial because such factors as differences in skills of the 
population in different places, differences in their ownership of factors of pro- 
duction, and preferences for living in certain areas might result in persistent 
income differences. But, do differences in wage levels more or less automatically 
set in motion forces that would result in their eventual elimination? Here the 
answer has to be Yes and No. To the extent that workers migrate over time 
from low-wage areas to higher-wage areas, wage levels in the different regions 
will tend to converge. New capital investment in low-wage areas will also tend 
to equalize wages. 

However, there are several reasons for qualifying this answer. First, the 
adjustments called for are not of the once-and-for-all variety; rather, we can 
expect that, as economic growth proceeds, adjustments of a continuing nature 
will have to be made. Take the example of agriculture. On the demand side 

4 This statement does not mean that all persons should leave agriculture or that some 
persons should leave agriculture from all places. Rather, it means that to achieve economic 
efficiency, some persons should leave agriculture from some places. 



102 / A Framework for Analysis 

for labor, the fact that in our economy the income elasticity of demand for most 
farm products is less than unity means that, as per capita incomes increase, a 
smaller proportion of national output will be devoted to farm products (at given 
relative prices). Likewise, continuing technological advance implies that a given 
output can be produced with less labor and other factors of production. Hence, 
it seems likely that growth in income at the national level will call for continued 
out-migration of labor from agriculture. On the side of capital, the analysis of 
location factors and input-output access makes it apparent that the level of wages 
is one but only one factor influencing new investment decisions. For certain 
industries, traditionally characterized as "labor-oriented," wage costs are the 
most significant factors affecting location. The cotton-textiles industry is per- 
haps the best example; others are the apparel and shoe industries. In recent 
years these industries have accounted for a declining share of total national 
employment. Of course, differences in wage levels might influence the location 
of new capacity in other industries as well. But for many industries, require- 
ments with regard to natural resources, markets, and transportation are the 
primary locational considerations. Unless the area could supply these require- 
ments, therefore, a given wage differential might not be sufficient to attract a 
significant in-migration of capital. 5 Finally, imperfections in the capital market 
might prevent the location of new industry in low-income areas. 

On the side of population, there are well-known institutional and behavioral 
forces which prevent workers and their families from migrating from low-wage 
areas at a rate sufficiently rapid to bring about an equalization of wages over 
relatively short periods of time. 

Under certain circumstances, however, the forces making for convergence 
can and do function with great effectiveness. Regional per capita income levels, 
taken as a whole, have converged on the national average during the period for 
which statistics are available — the years since 1880. More telling, however, is 
the experience of the various (multi-state) regions with relative growth in popu- 
lation and income, an experience which in a few cases is quite dramatic in 
highlighting the forces for convergence. 

The record which we have brought together reveals that the various regions 
have tended to "fill up" during the periods when labor was scarce in relation 
to job opportunities and wages were relatively high, as were also returns to 
capital, and that this "filling up" process tended to continue until wages and 
per capita income were brought up roughly to the national average and then 
to slow down or stop. For example, the Plains region's share of total national 
population rose from 1870 to a peak around 1890 and thereafter fell off quite 
rapidly. Its share of total personal income reached its peak a decade later, 
about 1900, and then declined sharply, somewhat more sharply than the popula- 

e The presence of "ghost towns" in the West provides mute testimony on this point. As 
mines gave out, wages in these areas no doubt fell relative to those in other areas. Since 
these mining areas offered almost no advantage to other types of production, the solution 
was for an outmovement of persons. Likewise today, industrial development has helped to 
alleviate the problems of some low income farming areas, but not all such areas offer 
sufficient advantage to new industry. 



Factors Influencing Per Capita Incomes / 103 

tion share. In the Mountain region, the rate of population growth was higher 
than that of the nation as a whole throughout the period 1870-1920. It was 
particularly high during the decades 1870-1890, slackened during the last decade 
of the century, but picked up again after 1900, though it failed to regain the 
rate of the eighties. In spite of these fluctuations, the region's share of total 
population continued to climb and only reached its peak about 1920 (at which 
time the Mountain region was still the most sparsely populated region in the 
country). The growth of total income had meanwhile not sufficed to raise the 
region's share of the national personal income total beyond the level of 1900; 
in fact, it had declined by some 2% by 1920. Income per capita in the Moun- 
tain region, which in the years around 1880-1900 had been among the highest 
of any region, converged downward rapidly toward the national average after 
1910, and after 1920 dipped somewhat below the average, where it remains in 
spite of some relative recovery after 1930. 

While the history of each region has unique features, the continuing play of 
the forces making for an adjustment between capital and population growth 
and movements can be noted. However, certain tendencies can be seen which 
help greatly to explain why the adjustments actually emerging do not provide 
for full covergence in regional per capita income. One of the most obvious, as 
well as one of the most powerful, of these tendencies is the greater ease of 
"filling in" when opportunities for capital and labor are unusually good than 
"emptying out" when opportunities (and wages) on the whole decline below the 
national average. The problems of liquidating capital investment once it is 
made, the difficulties of shifting to new jobs in other parts of the country for 
older persons, the attachment to home, community, and friends — these and 
similar factors are deterents to out-migration. The nature of these factors also 
helps explain why the "emptying out" process is particularly difficult in farm 
sections. Possibly most important of all, the birth rate is relatively high in low- 
income farm areas and even a high level of out-migration can be neutralized, 
at least in part, by a rapid natural increase in population. And yet, of course, 
farmers have left the soil in great numbers. The problem is essentially one of 
rate of net increase or decrease in population as against the rate of new capital 
formation in the region. 

As long as new investment decisions are based only partly on regional differ- 
entials in wage rates, and as long as only certain areas have the necessary ad- 
vantages in input-output access to attract a large inflow of capital, the critical 
factor for many of the low-wage low-income areas will remain the rate of out- 
migration. Any forces which pull industries away from locating in the areas 
which provide the best over-all situation for their particular requirements can 
do so only at the expense of the rate of increase in national product and average 
per capita income for the country as a whole. 




Summary 



Our examination of the many factors that influence the relative rate of increase 
in the volume of regional economic activities and per capita income levels has 
produced a number of quite significant YES, BUT propositions. These can be 
summarized as follows: 

1. Economic expansion is in general associated with a relatively high rate 
of increase in employment in manufacturing and services. Thus, growth of 
total income and population has been associated in a general way with declines 
in the relative importance of agricultural employment and increases in the 
relative importance of manufacturing employment, both for the nation as a 
whole and for almost every state separately. BUT there has been little relation 
between differential change, by states, in total income and population (i.e., in 
"volume") and the changing relative importance of different industries. 

2. "Growth" industries — those that are expanding in employment or value- 
added at a rate exceeding the average for all industries — favorably influence 
growth in the volume of economic activities within a region. BUT a region may 
grow by gathering in a greater and greater proportion of the slower-growth 
industries. Also, industry aggregates include a variety of industrial sub-cate- 
gories some of which are expanding more than others. Regions may experience 
growth even when they specialize in those industrial activities (such as agri- 
culture or mining) which may be on the decline on the whole. As a matter of 
fact, regions can be somewhat like individual firms. Just as some farmers, or 
coal mining firms, or textile or shoe manufacturing firms tend to make ex- 
tremely attractive profits and increase their output in situations where competing 
firms are having serious trouble, so there are farming and mining and textile 
areas which, by intensive production and the growth of service activities (which 
grow because of the generally high income levels) can experience growth when 
other areas with similar kinds of specialization are declining. Regions that are 
worried about decline can learn some useful lessons from the flourishing firms 
in generally declining industries. 

3. It is true that, no matter how the various industrial groups are aggregated 
or designated, certain industries are more conducive to regional growth than 
others. BUT not all regions have the relative advantages in regard to input- 
output access that make it possible for them to attract such industries; many 
can expect to grow only slowly on the basis of the industries for which they 

104 / 



Summary / 105 

do have special advantages. (One way of putting this is that not every area 
can hope to have the nice, clean, and fast-growing electronic and research 
industries.) Looked at in terms of relative advantages in resources, markets, 
human skills, amenities, climate, transport facilities and cost, and the rest of it, 
some areas can hope to grow only by attracting labor-intensive industries; 
others, by attracting certain kinds of processing industries exploiting relatively 
unexploited natural resources; some may have special advantages for certain 
types of assembly operations; still others for relatively intensive recreation 
activities, and so on. The attraction of industry is a competitive matter. There 
are many things that a region can do to enhance its locational advantages, par- 
ticularly with regard to human resources and social overhead facilities. But 
many features of nature and position within the nation are "givens." Our 
analysis suggests that a realistic appraisal of a region's relative advantages and 
disadvantages with regard to input-output access is an essential starting point 
for an understanding of its growth potential, as well as of its past growth. 

4. The growth of export industries within a region — those that tap the 
national and/or a broad regional market — strongly influence the expansion in 
the volume of economic activities. BUT the multiplier effect varies greatly with 
different industries, and the expansion of residentiary industries and services 
is not directly related to the growth of exports; the pattern of expansion of the 
residentiary industries and services within a given region is itself significant 
for differential growth as among the various regions. The direct and indirect 
impact on local income and employment of various export industries covers 
a full range from very great to relatively little; this is true of both the linked- 
industry effect (the expansion of business services, the local manufacture of 
parts and equipment, etc.), and the effect on increased consumer expenditures. 
Also, the extent of residentiary industries tends to increase with the growth of 
regional markets. As such markets grow in size with the growth of population 
and income, more and more industries find that they can achieve adequate 
scale economies while producing in every part of the country. This would seem 
to be particularly true of a wide variety of assembly operations and banking 
and financial activities, but many other types of activities tend to become almost 
ubiquitous as well, as far as the broad regions are concerned. 

5. A high capital-labor ratio (as will be shown in Part V) is necessary in 
providing a high level of wages within a given region, and different industries 
at any point in time vary greatly in the average pattern of factor proportions 
employed, so that wage level differentials are to be expected on this score. 
BUT there does not seem to be any inherent reason why any industry cannot 
increase the capital-labor ratio when adjustments in relative prices are conducive 
to such an increase. Thus, for example, certain types of farming have had 
rapid and large increases in labor productivity and the areas where these types 
of farming are located enjoy relatively high levels of per capita income. 

6. Every region in the country cannot hope to experience rapid increases in 
the volume of economic activities and in population. BUT every region can 
hope to enjoy a high and rising level of per capita income (as long as the 
nation's output and productivity increases) if it is willing to face up to the 



106 / A Framework for Analysis 

need for a relative "emptying out" when the over-all situation with regard to 
relative advantages among regions calls for it. 

These YES, BUT propositions add up to the following: All the regions of 
the United States are subject to the pervading influence of the national economy. 
At the same time, they are characterized by tremendous variety — in their phy- 
sical features, their closeness to national market centers, their existing plant 
and equipment, their population densities, their industrial structures, and their 
economic histories. Industries making their locational choices, given their spe- 
cial requirements, will inevitably find certain regions more attractive than 
others. But these requirements change over time, so that industrial shifts are an 
integral feature of a dynamic and basically rational economy. Thus, some 
regional economies will always be expanding more rapidly than others — essen- 
tially, on the basis of their relative advantages with regard to input-output 
access within the total spectrum of industries — old and new. Yet, as pointed 
out above, while not every region can expect equally rapid growth in the 
volume of economic activities, each can enjoy a general level of living not 
far away from the level of the nation as a whole. Policy at both the national 
and the local levels might well be directed to the achievement of such an 
objective. 



Part 

in 



Regional Economic Development 
1870-1950 



10 Regional differentiation in the 

American economy before 1870 



The variety of possible growth sequences in a developing economy is well 
illustrated by differing regional experiences in the United States after 1870. 
The modern continent-wide economy has its roots in the small and relatively 
isolated settlements planted along the Atlantic seaboard by European (mostly 
British) families and investors during the seventeenth century. For long these 
thinly populated settlements, clustered around some port community, were 
largely unconnected with each other but maintained vital ties with northern 
Europe and the West Indies. Only in the late seventeenth century did their 
common connections assume any regular or significant pattern. 

By 1700 there were perhaps 300,000 people unevenly distributed along the 
maritime fringe; few had made large fortunes and their collective economic 
achievements were modest indeed. The great expectations of early investors and 
landholders had been mostly disappointed and, with minor exceptions, interest 
in the growth potential of mainland settlements north of the Chesapeake area 
had sharply declined. But within the next six decades population was to in- 
crease fivefold and a notable differentiation of productive activities and levels 
of well-being was to appear. Relations with the wider Atlantic community prob- 
ably played a more decisive role in this transformation than those among the 
mainland communities themselves, but, except for the plantation economies of 
the Chesapeake and Carolina regions, progress was as much a consequence of 
local efforts as of aid from outside. 1 

In all the colonial settlements, labor and credit were in short supply; land 
was the only factor in abundance. From the outset, therefore, development was 
closely governed by local resource availabilities, and market outlets were 
necessarily confined within narrow reaches. Where great commercial crops — 
tobacco, rice, or sugar — could not be produced, where forests did not yield 
naval stores or furs, primary production remained relatively undifferentiated. 

1 C. Bridenbaugh, Myths and Realities: Societies of the Colonial South (Baton Rouge: 
Louisiana State University Press, 1952) ; idem. Cities in the Wilderness: The First Cen- 
tury of Urban Life in America, 1625-1742 (New York: Ronald Press Co., 1938) ; E. B. 
Greene and V. D. Harrington, American Population before the Federal Census of 1790 
(New York: Columbia University Press, 1932), p. 4. 

/ 109 



110 / Regional Economic Development: 1870-1950 

By the mid-eighteenth century, to be sure, cattle, wheat, and smaller grains 
were raised for urban markets by cultivators in both the Middle Atlantic and 
the Southern colonies; wheat and dairy products were also obtained from the 
thickly settled river valleys of New England. Manufactures (mostly handicrafts) 
and trade were likewise based upon local materials extracted from forests, 
fields, and seas; the New England fisheries and carrying activities were a major 
source of income for many coastal communities. But the availability of land 
at this time probably aggravated the shortage of labor and deterred the ac- 
cumulation of venture capital; numerous immigrants who might have done 
well as artisans or laborers chose self-employment at a lower level of living on 
the land rather than work for a master once their terms of indenture had ex- 
pired. 2 Under such conditions much of the needed social overhead was not 
forthcoming and the bulk of commerce was restricted to local exchange and the 
coastal trade. Only credit and the complex geometry of the Caribbean connec- 
tion enabled the northern colonies to obtain their modest provision of consumer 
goods from the merchants of Europe. Even at the close of the colonial period 
the coastal trade was probably no more than a third of the combined ocean 
trade with Britain and the West Indies in value, although it greatly exceeded 
the latter in volume. 3 

No doubt the low rate of capital formation was in some small part the out- 
come of mercantile restriction in the interest of the mother country, but it was 
also a consequence of the fact that credit was always scarce and borrowing 
expensive. In any case, the rich endowments of forests, fisheries, and farm 
lands made for an unquestionable advantage in sectors which required only 
small fixed investment in tools and equipment. Yet if all colonial communities 
had depended exclusively on the export of services or natural resources, it is 
likely that few would have proved viable. As it was, the plantation economy 
of the South was for almost two centuries the richest section on the mainland; 
its soil and climate combined with slave labor to give it a marked advantage in 
the cultivation of certain staples which were in regular, if somewhat erratic, 
demand throughout the growing Atlantic economy. 4 For reasons of soil and 
climate, New England was by far the poorest agricultural section, and by the 
mid-eighteenth century it was already a net importer of certain foods and 

2 R. M. Tryon, Household Manufacturers in the United States, 1640-1860 (Chicago: The 
University of Chicago Press, 1917) ; R. G. Albion, Forests and Sea Power: The Timber 
Problem of Royal Navy, 1652-1862 (Cambridge: Harvard University Press, 1926) ; H. A. 
Innis, The Cod Fisheries: The History of an International Economy (New Haven: Yale 
University Press, 1940), pp. 160-62; R. B. Morris, Government and Labor in Early 
America (New York: Columbia University Press, 1946). 

3 J. S. Homans, Historical and Statistical Account of the Foreign Commerce of the United 
States (1857), pp. 6-7; R. Pares, Yankees and Creoles: The Trade Between North America 
and the West Indies before the American Revolution (London or New York: Longmans, 
Green and Co., 1956) . 

4 G. D. Ramsay, English Overseas Trade during the Centuries of Emergence (London: 
Macmillan, 1957), pp. 207-39. C. Bridenbaugh, op. cit., p. 14, emphasizes that by 1765 
the Chesapeake Tidewater economy faced bankruptcy and that only the new Piedmont 
lands enjoyed prosperity thereafter. 



Regional Differentiation in American Economy before 1870 / 111 

fibers. Thereafter its growth was assured only by the development of handi- 
crafts, distilleries, shipbuilding (built on forest resources), and other services 
to maritime commerce, in all of which it came to excel. Resource availabilities 
in the Middle Colonies had meanwhile made for a better balance between nas- 
cent manufactures and husbandry, between the more differentiated activities of 
town and country, and had fostered a progressive division of labor in support 
of the rising population. 5 Eventually, the adaptation of water power in the 
river valleys of the Middle Atlantic and New England colonies permitted local 
concentrations of mills and furnaces, though here, too, larger-scale investments 
and operations had to await the growth of closer linkages among the separate 
settlements and a more intensive exploitation of their common market potential. 6 

The disparity between conditions of life in the Old World and the New ex- 
cited the comment of observers long before the English colonies entered upon 
their separate political career and, if anything, the awareness of difference 
had increased as the nineteenth century progressed. As Tocqueville remarked 
in 1835, fortune had indeed proferred Americans "an immense booty." Not 
that this booty was ever there just for the taking; a special kind of human 
resourcefulness was needed in order to realize the bountifulness of nature. It 
was never enough simply to take up new lands in the West; this may have 
postponed the need to face the Malthusian challenge, but in itself "land" did 
not provide an answer. If Americans had simply looked to the land for sub- 
sistence, they might have turned into a nation of peasants; as it was, most cul- 
tivators, large and small, hoped to raise a sizable crop surplus for market, while 
growing numbers turned to trade and eventually to manufactures. 7 

Between 1790 and 1810 the number of cities with more than 8,000 inhabi- 
tants increased from six to eleven. Communities of 2,500 or over rose by 43% 
during these two decades. In the decade after 1810, there was a relative decline 
in urban population owing to the interruption of commerce by the War of 1812 
and to the taking up of new land in the West. As late as 1820 only six cities 
had populations exceeding 20,000. New York and Philadelphia, however, both 
exceeded 100,000, and Baltimore had expanded to 62,000. None of these urban 
centers was yet a city of "factories" in the sense of Manchester or Lyons in 

6 By the mid-eighteenth century English producers complained that people in the Middle 
Colonies were "sliding into the manufactures proper to the mother country."; E. E. Lam- 
pard, "History of Cities in the Economically Advanced Areas," Economic Development 
and Cultural Change, Vol. 3 (January 1955), pp. 116-17. 

6 The potential greatness of the American economy was nevertheless apparent by the 
last quarter of the eighteenth century. Already in 1776 Adam Smith foretold that Ameri- 
cans would create "an extensive empire . . . one of the greatest and most formidable 
that ever was in the world." He saw America surpassing the wealth of England in little 
more than a hundred years. Wealth of Nations, ed., E. Cannan (New York: Random 
House, Inc., 1937), pp. 588 and 590. Also, D. M. Potter, People of Plenty: Economic 
Abundance and the American Character (Chicago: The University of Chicago Press, 1954), 
pp. 78-86. 

7 For the continuing dependence of American expansion on the Atlantic trading area, 
see Frank Thistlethwaite, "Commercial America," in H. C. Allen and C. P. Hill, eds., 
British Essays in American History (London: Edward Arnold (Publishers) Ltd., 1957), 
pp. 102-105. 



112 / Regional Economic Development: 1870-1950 

the Old World. The expansion of all, with the possible exception of Philadel- 
phia, was essentially a commercial-service growth due to the swelling of trade 
as the resources of the interior became available for export. New York had 
been especially successful in attracting trade from the back country. The com- 
pletion of the Erie Canal in the middle twenties clinched its supremacy and 
inaugurated a nation-wide mania for internal improvements. 8 

Among rivals for third place after New York and Philadelphia, Baltimore 
had grown by nearly 80% in the decade of the second war with England, 
Boston by little more than 30%. Baltimore's geographical position and mer- 
cantile enterprise combined to give it a notable advantage in the growing com- 
merce of the North Atlantic seaboard region. In 1328 its merchants began 
construction of the Baltimore and Ohio Railroad with which to challenge even 
the supremacy of New York and Philadelphia. Boston, on the other hand, was 
losing ground in this clash of urban imperialism. Though its fleet was un- 
matched and its growth considerable, its geographic location in relation to the 
main thrust of internal development and the paucity of resources in its own 
back-country combined to threaten its position as the chief entrepot north of 
the Hudson. After 1815 the Boston merchants and their associates set out to 
corner the entire New England trade, and their success can be measured in the 
relative decline of most other communities in the region. The Boston merchants, 
in fact, were forced to create business, and this they did by promoting an in- 
dustrial revolution in Massachusetts. The impetus to the integration of spe- 
cialized manufacturing processes in urban factories came not so much from 
the semi-rural centers of handicrafts in Rhode Island or Massachusetts as from 
the embattled commercial metropolis. 9 By importing raw material from South- 
ern states and combining it with surplus labor from the countryside at strategic 

s Robert Albion suggests that New York had more than positional advantage in the 
struggle for commerce; its business leaders had established a very successful auction system 
for disposing of imports, had a growing coastwise trade, especially with parts of the South, 
and had regular and reliable transocean packets; The Rise of the Port of New York, 
1815-1860 (New York: Charles Scribner's Sons, 1939), p. 13. By the early 19th century 
Philadelphia already had a diversified base of manufactures: textiles, hats, grain and saw 
mill products, hosiery, paper, printing, sugar, iron foundry and tool-making products, 
carpets, carriages, stoves, and ships; it was also a center of rationalist thought and scien- 
tific inquiry. See E. E. Lampard, loc. cit., p. 117. 

9 The old tradition of semi-rural manufactures in both New England and the Middle 
Atlantic states had helped "industrialize" a section of the labor force and had nurtured a 
familiarity with machine processes. Even before 1820 some $600,000 of mercantile capital 
went into the so-called "Waltham Plan" to finance the improved power loom and to inte- 
grate the specialized processes of cotton cloth manufacture in single plants under unified 
managements, see G. R. Taylor, The Transportation Revolution, 1815-1860 (New York: 
Rinehart, 1951), pp. 207-249; C. F. Ware, The Early New England Cotton Manufacture 
(Boston: Houghton Mifflin Co., 1931) ; V. Shlakman, "Economic History of a Factory 
Town: A Study of Chicopee, Massachusetts," Smith College Studies in History, Vol. 20, 
(1934-35) ; R. K. Lamb, "Entrepreneurship and Community Development," Explorations 
in Economics . . . contributed in Honor of F. W. Taussig (New York: McGraw-Hill 
Book Co., 1936), pp. 526-534. 



Regional Differentiation in American Economy before 1870 / 113 

points along the natural waterways, merchant capitalists of Boston accelerated 
a social process which lifted southern New England over the initial threshhold 
to rapid economic growth. Other communities followed suit and the territorial 
division of labor between town and country and between region and region 
was thus enabled to exert its beneficent sway. 

By 1840, one historian states, "It would have been difficult to find 50 out of 
the 479 townships in southern New England which did not have at least one 
manufacturing village clustered around a cotton or a woolen mill, an iron 
furnace, a chair factory or a carriage shop, or some other representative ... of 
manufacturing which had grown up in haphazard fashion in every part of the 
three states. 10 The same process was also at work by 1840 in parts of the Mid- 
dle Atlantic region and in the Ohio Valley. 11 

Estimates of per capita income originating in commodity production and 
commerce by states in 1840 provide some measure of this early differentiation. 
The nation-wide average per capita income from these sources in that year was 
around $65. New England, more especially southern New England, had by 
1840 surpassed all other states and territories except Louisiana (whose small 
population dominated the growing trade of the Mississippi system). The high- 
est per capita figure for any state was that of Rhode Island, $118. Louisiana 
was next with $113; but then came Massachusetts with $107 and Connecticut 
with $91. Elsewhere in the expanding South only Mississippi ($84), Arkansas 
($68), and Florida ($69) exceeded the national average. New Jersey ($83), 
New York ($80), and Pennsylvania ($75) in the Middle Atlantic region and 
the Wisconsin Territory ($80) in the Great Lakes region were also well above 
the national average. Nonagricultural income per worker excluding commerce 
was nearly 122% of the national average of $338.9 in the West South Central 
states and 102.5% in the Middle Atlantic states; in New England it was almost 
138%. 12 (See Table 16.) These estimates give some indication of the extent 
to which New England was transforming its economic base away from agricul- 
ture and commerce before the mid-century. 

Only during the second and third decades of the nineteenth century, however, 
as a result of the impact of technological advances in transportation on the 
pattern of resource availabilities, did a dynamic system of regional economic 

10 P. W. Bidwell, "The Agricultural Revolution in New England," American Historical 
Review, Vol. 26 (1921), p. 686. In spite of New England's demand for cotton, the value of 
raw cotton rose from less than 40% in 1821 to nearly 60% of all U. S. exports in 1851. 

"In 1840 only New England and the Middle Atlantic states had greater concentrations 
of nonagricultural workers than the nation as a whole, 38.4% and 32.1% of their respective 
labor forces compared with 20.8% for the nation. Greatest concentrations of agricultural 
workers were to be found in the labor forces of the East South Central states (92.1%) 
and the South Atlantic states (89.9%). See Table 16. 

12 R. A. Easterlin, "Interregional Differences in Per Capita Income, Population, and Total 
Income, U. S., 1840-1950," National Bureau of Economic Research, Conference on Research 
in Income and Wealth, September 4-5, 1957 (mimeo.), Table A-3. Easterlin's estimates are 
based principally on the 1852 edition of Ezra C. Seaman, Essays on the Progress of 
Nations (New York: Charles Scribner, 1852). 



114 / Regional Economic Development: 1870-1950 



H3 




O 




S 




s 




O 




o 




c 




• r-l 




WO 




c 












cc 




c 








bD 








-— 




o 




cp 




S 


O 


o 


CO 


1 


1— 1 








c 


H3 


# o 




"5b 

03 




Sh 




>> 


'- 


^5 


o 




MH 


CO*" 


>- 


C15 


o 


"o3 




cjq 


F— * 


H3 


C 


CD 


^O 


"8 


J2 


P 






3 


o 


PL, 

o 


JG 


Ph 


.5 




C 


vo 


^O 


"*H 


o 


<U 


3 




O 



a 1 

o 



CD 

s 

. o 

H 

d a 

o 
S5 



. CD 

•8 £ 

& 8 
< .a 



3 I £ § 

a fc B -o *i3 

° § s S 

£ 3 1 I 

P-i C G ft 



2 1 



b S3 

•■S o 

ft **-i T3 ".3 

ed o cj 

O co - 



H 



^ bfi 
<< cd 



3 '- , 

ft .2 

o *-> 
Cm 



v© <-0 CO On vo vq ^. 

to tsl On t>* i-i CM* «T5 

vo ^ co cm co lo a 

^ fO CN N CN N ^ 



O ^ ^ ^ 00 VO ^. 

io cn co M* ej d 2 

« VO Tj N ^? t^ © 

•^ ^i M «5 « CO 0\ 



N ^ to W CN CC C 

r-H >/J n o\ oo' in oo 



""* CO Tf T# VO t^; <N 
^5 CNj o> CN ON O CN) 

N ^O CO <* t? lO N 



cnj o i— i i— i vo i-h vq 

CO VO VO 1-4 ■* LO Crj 
CO K r? " 



lO lO lOO 



Tf 0\ H N M pq . 
OV f- O O t »0 CN 
CO CO vo 



v^ >^~ uj w>i 

i— I OV f~- r— I 



$ 



VO CM O 

"vT r— I 



lO ^J 1 lO N ^ lO O 
CO l> CM CM £2 h- LO 
CM O C\ r* CO LO 'sf 



bfc 



m 



-o d 

bD <! ed 

C CD r^ M 

P S3 S O 





c 


o 


CO 




U 


d 




ed 


-g 






<j 


"3 

o 




en 


M 








3 


03 


o 


cd 


V) 


W 



C/2 co 

pi 

.. <M 

* J* 

s ^ 

s s 



O 



«, * 3 

•rt I— I * 



Cd J 

Jg 

ft - 

m 

1 & 



8 J 



a3 < 



S 8 

d c 






el 

o 
en 



Regional Differentiation in American Economy before 1870 / 115 

interdependence emerge on the mainland of North America. 13 Corn and salt 
pork from the Ohio Valley, grain from the prairies of Ohio and Illinois, lumber 
and cotton from Tennessee and Mississippi, were gathered in from country 
points by the merchants of Pittsburgh, Cincinnati, Louisville, and St. Louis 
and shipped aboard river steamboats to New Orleans, there to be transferred 
to coastal vessels bound for the Atlantic seaboard or to ocean freighters for 
the ports of western Europe. By the early 1840's the volume of grain shipped 
east via the Great Lakes exceeded that shipped down the Mississippi; and even 
though the commerce of New Orleans continued to grow, the Atlantic ports 
began to dominate the business of the interior. 

These were by no means autonomous developments; westward expansion and 
regional differentiation were closely linked to general expansion of the greater 
Atlantic trade area, but by the mid-century the rise of the textile industry in 
New England and the more gradual and diversified growth of manufactures in 
the Middle Atlantic states marked the beginnings of economic "independence." 
By that date a quarter of the raw cotton crop was being diverted to meet Amer- 
ican needs and, though the balance of international trade remained generally 
passive until after 1870, dependence on foreign manufactures slowly declined, 
more rapidly after 1860. The introduction of the high tariff during the Civil 
War years gave added impetus to industrial development, accelerated differen- 
tiation in the Northeast, and led directly to the expansion of the domestic 
market for American manufactures without depriving the economy of benefits 
it derived from the export of primary produce. 

During the third quarter of the century, therefore, primary producers in the 
Trans- Appalachian West were adjusting to the demands of two industrial cen- 
ters: Western Europe and the Northeastern states. Some found opportunities 
even closer at hand and pondered the question whether greater enrichment lay 
in developing local markets or in reducing the cost of moving surpluses to 
Southern and Eastern entrepots. Cotton growers were never wholly reconciled 
to their dependence on the milltowns of Lancashire and New England; wheat 
growers and cattlemen took advantage of local outlets, provisioning newcomers, 
army posts, mining and lumber camps within the radius of a two days' journey. 
The sequence of early land entries in parts of the Old Northwest indicate that 
settlers and speculators had an eye for local market potential as well as for the 
general accessibility or natural properties of an area; town sites and farm 
lands were often developed together. 

The historic predicament of the primary producer in the West arose from 
the fact that his locality usually failed to grow (i.e., differentiate) at a rate 
sufficient to absorb incremental outputs made possible by further inputs of 

13 To surpass the average levels of material existence achieved in Europe, the U. S. did 
not require the unified economy or the centralized political institutions of its counterparts 
across the ocean. Its vast land area, its diversity of natural endowments, the variety of 
institutions and jurisdictions under which its peoples lived had retarded "nationalizing" 
tendencies without obvious cost to the general welfare. It was a virtual customs and pay- 
ments union; so far it had enjoyed the advantages of union without completing the politi- 
cal struggle for unification. 



116 / Regional Economic Development: 1870-1950 

labor and capital on virgin soils, thereby heightening his dependence on extra- 
local markets and a train of intermediaries in transportation and wholesaling. 
It was for this reason that the early grain growers of Wisconsin, for example, 
seized upon the railroad as the sure way to "bring New York to our doors." 14 
Only improved transportation could bring the primary producer closer to 
his market and at the same time facilitate the growth of more diversified activi- 
ties nearby. The development of competitive water routes had fostered differ- 
entiation in the interior, but all water courses were closely circumscribed by 



Table 17. Percentage distribution of United States population in the "West,' 
1340 and 1860 



Region 1840 1860 

Jo in "West"* 37.5 49.2 



Great Lakes 17.1 22.0 

South Central (west of Appalachians) 17.9 16.5 

Plains 2.5 6.9 

Southwest — 2.2 

Mountain — 0.2 

Far West — 1.4 

*U. S. total population was 17,020,000 in 1840 and 31,400,000 in 1860. 

Source: R. A. Easterlin, loc. cit. Tables A-2 and A-3, see Table 16; 17th U. S. Census, 
1950, see supra, Table 1, note. 



the capacity of vessels, handling facilities, and by the highly seasonal character 
of inland navigation. 15 Nevertheless, the availability of cheaper western food 
tended to depress food production in the South and Northeast, forcing the one 
region to intensify its commitment to the cotton staple and the other to com- 
merce and manufactures. Only the iron railroad could at once free western 
activities from the paralytic grip of winter and emancipate developments in 
coastal sections from the directions imposed by primary river flow. 

. . . when the railroad penetrates our borders, and gives us at all times access to the 
Atlantic Seaboard, what more can the farmer ask or need, but well directed industry, to 
crown his efforts?" Wisconsin State Agric. Society, Transactions, (Madison: Beriah Brown, 
State Printer, 1853), Vol. 2, (1852), p. 104; ibid., Vol. 3, (1853), pp. 115-122. 

15 By 1820 some sixty steamboats had reinforced the flat-boat fleet on the Mississippi sys- 
tem; by 1840 their number exceeded four hundred, and by 1860 it had reached a thousand. 
Growth of steam navigation on the Great Lakes was somewhat slower, but by 1840 the 
quantity of grain shipped via the Great Lakes probably exceeded that moving down the 
Mississippi. In 1820 receipts of Western produce formed 58% of commodities reaching 
New Orleans, and by 1860 only about 23% of a larger total. See L. C. Hunter, Steamboats 
on the Western Rivers (Cambridge: Harvard University Press, 1949) ; G. R. Taylor, op. cit., 
pp. 56-73. 



Regional Differentiation in American Economy before 1870 / 117 

The strongly agrarian character attributed to the westward movement by 
most historians and the emphasis placed by others on western railway develop- 
ment during the mid-fifties have tended to obscure the influence exerted by a 
third factor — urbanization. The growth of cities was by no means confined 
to the more industrialized coastal sections. In every new country the growth 
of central places is an essential concomitant of economic development, since no 
efficient organization of settlement can proceed very far without creating non- 
agricultural communities which carry on the vital functions of collecting and 
distributing articles of trade. From the early nineteenth century, if not before, 
certain places in the interior — the "valley centers" and "lake centers" — had 
served as entrepots for western commerce. 16 Around their mercantile core had 
grown up a variety of processing and servicing activities such that by 1860, 
when more than half the total population of thirty-one millions lived west of 
the Appalachians (Table 17) and more than 14% west of the Mississippi (ex- 
cluding Louisiana and Arkansas), seven western cities had populations exceed- 
ing 50,000 (Table 18), with sizable proportions occupied in manufactures. By 
this time, also, over 23% of all employment in manufactures was located west 
of the Appalachian divide, more than half of it in the Great Lakes states. 

It is no exaggeration to say that Western promoters had sited, and in some 
cases developed, their urban real estate almost before the farmers had brought 
in the first commercial harvest. During the second quarter of the nineteenth 
century almost every crossroads village in the West had postured as the future 
"Queen City" of its region, serving the vast primary producing areas around 
it and eventually surpassing the great centers of the Eastern seaboard. They 
vied with each other in a fabulous struggle for "internal improvements," to 
build plankroads to link up rivers with canals and later to join river systems 
to the Great Lakes with railways. Though on a much smaller scale than the 
historic rivalry between Boston, New York, Philadelphia, and Baltimore, the 
feeling was no less intense and the rhetoric no less rotund. The struggle be- 
tween Milwaukee, Chicago, and St. Louis to dominate the territory where the 
Great Lakes touch the Mississippi system is but the most famous and best-doc- 
umented of these urban rivalries; its essence was repeated in microcosm 
through every section of the interior. 17 

16 R. C. Wade, "Urban Life in Western America, 1790-1830," American Historical Review, 
Vol. 66, (1958) ; C. E. Reiser, Pittsburgh's Commercial Development, 1800-1850, (Harris- 
burg: Pennsylvania Historical and Museum Commission, 1951). 

"The creative force of urban development can be traced in the public press and in the 
reports of local chambers of commerce and city clubs. By the fifties, spokesmen for 
Western cities had realized that their growth potential was limited so long as it was based 
exclusively on commerce. The editor of the Cleveland Leader declared: ". . . no thinking 
man with capital will stop here when we have only commerce to sustain us. A manufac- 
turing town gives a man full scope for his ambitions." His paper advocated popular sub- 
scriptions to factory enterprise, reductions in the prices of real estate, protection of "home 
manufactures." A similar stand was taken 1863 by the new Milwaukee Manufacturers' 
Association, and the city's Sentinel insisted: "Commerce alone can never give us permanent 
prosperity." A decade later the Cleveland paper affirmed on some unstated authority 
that "a thousand dollars put into manufacturing does more to gather population than a 
million dollars put into trade." In short, business leaders in many Western cities had 



118 / Regional Economic Development: 1870-1950 

The scale of urban development is registered in the size of city populations. In 
1840 only New Orleans of the cities outside the East, exceeded 100,000 popula- 
tion and, strictly speaking, this was a maritime-colonial city like old Boston or 
Philadelphia rather than a true city of the interior. But by 1860 Cincinnati and 
St. Louis had populations of over 150,000, and Chicago had passed the 100,000 
mark (Table 18). Louisville and San Francisco numbered over 50,000, and 
half a dozen other cities had grown beyond 20,000. Within the next decade 



Table 18. Larger urban centers of the "West" in 1860, ranked in order of 
population engaged in manufactures 

Urban center Total population Percentage in manufactures 

Cincinnati 161,044 18.3 

Louisville 68,033 9.8 

Buffalo 81,129 6.9 

St. Louis 160,773 5.8 

Chicago 109,260 4.9 

New Orleans 168,675 3.0 

San Francisco 56,802 2.6 

Source: 8th U. S. Census, 1860, Mortality and Miscellaneous Statistics, p. xviii. 



seven cities in the interior had more than 100,000 inhabitants and, with the 
first trunk railroads linked to New England and New York, the main thrust 
of urban development had shifted from the Ohio Valley to the Great Lakes' 
shores. The five Lake cities — Buffalo, Cleveland, Detroit, Chicago, and Mil- 
waukee — had a combined population in 1870 that was sixteen times their com- 
bined population in 1840; in the same interval their states had increased their 
populations only threefold. 18 

When the character of American manufactures in the years before the Civil 
War is considered, there is every reason why urban-industrial centers should 
have emerged in the midst of the great resource-producing hinterland. The 
leading branches of manufacture in 1860 included many processing activities 
working on primary raw material. Where weight losses were substantial in 
any process, it was likely that the relatively high level of transfer costs would 
foster local or regional processing: lumber products, flour and grist-mill prod- 
realized what was apparent to their forebears in Boston more than half a century before, 
that commerce in raw materials was a wasting asset and that future growth "involved 
substituting the encouragement of manufacturing for an earlier emphasis on trade." 

See also W. W. Belcher, The Economic Rivalry between St. Louis and Chicago, 1850- 
1880 (New York: Columbia University Press, 1947) and Bayrd Still, Milwaukee, the History 
of a City (Madison: State Historical Society of Wisconsin, 1948). 

M B. Still, "Patterns of Mid-Nineteenth Century Urbanization in the Middle West," 
Mississippi Valley Historical Review, Vol. 28 (September 1941), pp. 187-206. 



Regional Differentiation in American Economy before 1870 / 119 

ucts, and some types of leather products were eminently adapted for western 
locations. Taken together they made up a sizable proportion of the total value 
of manufacturing output, if a somewhat smaller share of value added by manu- 
facture (Table 19). 



Table 19. Leading branches of manufacture in the United States, 1860 











Value added 


Rank 








Value of 


by 


by 








product 


manufacture 


value 




Item 


Employment 


(000's of $) 


(000's of %) 


added 


1. 


Flour & meal 


27,682 


248,580 


40,083 


4 


2. 


Cotton goods 


114,955 


107,338 


54,671 


1 


3. 


Lumber 


75,595 


104,928 


53,570 


2 


4. 


Boots & shoes 


123,026 


91,889 


49,161 


3 


5. 


Men's clothing 


114,800 


80,831 


36,681 


5 


6. 


Iron (cast, forged, rolled, 












wrought) 


48,975 


73,175 


35,689 


6 


7. 


Leather 


22,679 


67,306 


22,786 


9 


8. 


Woolen goods 


40,597 


60,685 


25,030 


8 


9. 


Liquors 


12,706 


56,589 


21,667 


10 


10. 


Machinery 


41,223 


52,010 


32,566 


7 



Source : 8th U. S. Census, 1860, Manufactures, pp. 733-42. 



Table 20 highlights the dominant place of the Middle Atlantic and New Eng- 
land regions in the nation's early manufactures, but it also points to the fact 
that by 1860 processing outside of these two regions was no longer negligible. 
Particularly it shows the beginnings of a rapid growth in manufactures which, 
in the period after 1870, would tie the Great Lakes region into the Northeastern 
industrial belt. Yet, taking the country as a whole, the combined value of three 
major farm crops (corn, wheat, and hay) in 1860 exceeded the total value 
added by manufacture, while capital invested in manufactures was less than 
one-sixth the value of all agricultural land and buildings. As yet there was 
only a beginning of the enormous industrial transformation that would be ex- 
perienced more or less in all parts of the country by the post-bellum generation. 

Finally, we conclude this brief review by considering possible relationships 
between differential levels of living and occupational structures toward the end 
of the period. Table 21 shows how the eight regions ranked in real personal 
income per capita in 1880 (the first date for these figures) and gives data on 
the urbanization of their populations and on the distribution of their labor 
forces among the agricultural, nonagricultural-resources, and manufacturing 
components of employment. The highest levels of per capita personal income 
were achieved in the as yet largely underdeveloped Far West and Mountain 
regions and, the next highest in the urban-industrial New England and Middle 



120 / Regional Economic Development: 1870-1950 



Table 20. 
and 1860 



Employment in manufactures in the United States, by region, 1850 



Region 



1850 



1860 



United States 



New England 

Middle Atlantic 

Great Lakes 

South (East of Appalachians) 

South (West of Appalachians) 

Plains 

Southwest 

Mountain 

Far West 



958,079 


1,311,246 


100.00% 


100.00% 


32.64 


29.88 


43.90 


41.66 


9.70 


12.09 


6.28 


5.45 


5.08 


4.36 


1.83 


2.30 


.12 


.34 


.01 


.03 


.44 


3.89 



Source: 9th U. S. Census, 1870, Vol. Ill, Table VIII (A) 



Table 21. Rank order of regions by personal income per capita, urbanization 
of population, and labor force industry components, 1880 





Personal 


Urban 










income 
per capita 


proportion 
of 


Labor force components 




% Agri- 


% Non-agri.* 


% Mfg.- 


Region 


(1929 dollars) 


population 


culture 


resource 


service 


United States 


$302.1 


28.2% 


49.39 


2.27 


48.34 


Far West 


638.2 


35.9(3) 


28.81(5) 


14.46(2) 


56.73(3) 


Mountain 


501.5 


25.0(5) 


24.00(7) 


27.35(1) 


48.65(4) 


New England 


425.0 


52.4(1) 


21.46(8) 


1.51(6) 


77.03(1) 


Middle Atlantic 


422.8 


49.8(2) 


24.02(6) 


2.91 (3) 


73.07(2) 


Great Lakes 


307.4 


27.5(4) 


50.75(4) 


1.67(4) 


47.58(5) 


Plains 


271.2 


18.2(6) 


61.36(3) 


1.23(7) 


37.41(6) 


Southwest 


182.7 


9.2(8) 


72.33(2) 


1.58(5) 


26.09(7) 


Southeast 


150.6 


9.4(7) 


75.36(1) 


0.64(8) 


24.00(8) 



Figures in italics are above the national average. 
* Forest and logging, mining, and fisheries. 
( ) Rank order. 

Source: Unpublished data by R. A. Easterlin; 17th U. S. Census, 1950, see supra, Table 
4, note; and Appendix Tables A1-A7. 



Atlantic regions; the lowest were in the rural-agricultural regions of the South- 
west and Southeast. In between the higher urban-industrial regions and the 
lower rural-agricultural regions were situated the Great Lakes and Plains regions 
(mostly eastern Plains states), with per capita income levels closest to the 
countrywide average. 



Regional Differentiation in American Economy before 1870 / 121 

Thus, by 1880, there were, broadly speaking, four different regional levels of 
living in the United States as measured by per capita personal income, and each 
was associated with a characteristic occupational structure of the regional labor 
force. In the first group were the sparsely populated nonagricultural-resource 
regions of the Far West and Mountain regions with high concentrations of their 
small labor forces employed in mining. The medium concentrations of industrial 
employment in these two regions were in services rather than manufactures, and 
both regions had relatively low concentrations of agricultural activities. The 
second group comprised the two densely populated regions of the North Atlantic 
seaboard, New England and the Middle Atlantic states. At this time, New Eng- 
land, with the higher concentration of manufacturers, seems to have had a 
slightly higher average level of per capita income than the Middle Atlantic 
region, but the difference was probably not significant. In the third group came 
two rapidly developing regions of industrial and agricultural potential, the Great 
Lakes and the Plains region, the former already more urban-industrialized and 
enjoying a higher average level of living than the still largely unsettled Plains. 
The fourth group, comprising the heavily populated Southeast and the partly 
settled Southwest, were the most agricultural and the least industrial regions in 
the country; as a consequence of their relative economic "backwardness," their 
personal income per capita was little more than half the national average. 19 

Such were the different regional levels of living in the United States at the 
end of the 1870's. On the face of it, some regions were relatively poorer in 1880 
than they had been in 1840 and some were relatively much richer; the range 
from the poorest to the richest was certainly much greater than it had been at 
the earlier date. That factors of production could not yet move "freely" through- 
out all parts of the continental territory, seems implicit in the extent of regional 
divergence in income which, if anything, had grown over previous decades. The 
remote undeveloped Mountain and Far West regions had achieved their excep- 
tional levels of per capita income because they were largely unsettled and pos- 
sessed the bulk of the nation's domestic supply of precious metals; wages in 
these outlying areas were usually much higher than in other parts of the country. 
The Plains and Southwest, meanwhile, had considerably larger populations but 
apparently lacked resource availabilities other than agriculture. On the other 
hand, the highly developed New England and Middle Atlantic regions had 
already achieved a rapid rate of urban-industrial growth and were shortly to be 
joined by the Great Lakes region. The Plains, however, failed to realize its 
earlier promise and, if it did not revert to the lower levels of the Southeast or 
Southwest, it had nevertheless declined relative to the urban-industrial regions 
by the close of the century. A similar fate overtook the Mountain region some- 
what later. The changing fortunes of the several regions as they joined in the 
burgeoning national economy after 1870 are the subject of the next chapter. 



19 There can be little doubt that the degree of Southern "backwardness" was largely the 
result of Civil War dislocation as well as relative "overpopulation" of the countryside. See 
Table 16 above for the 1840 figures by contrast. 



Jl / Changing regional structure of the 
U.S. economy, 1870-1910: 

(1) GROWTH AND SHIFT OF POPULATION 



The phase of American economic history dominated by the self-conscious striv- 
ing of rival seaboard communities to develop the resource potential of the 
interior ended during the third quarter of the nineteenth century and the period 
of "national" development was finally well under way. Between 1870 and 1910 
the economy was rapidly industrialized, a far-reaching social process which 
affected not merely manufactures but resource and service activities as well. A 
growing population, enlarged by millions of immigrants, surged out from the 
eastern half of the country into the greater West; a trickle of miners came east- 
ward across the Sierras into the Southwest and Mountain regions. Railroads, 
partly financed by European capital, tied the Atlantic to the Pacific and sent 
long fingers into remote regions to secure the untapped resources of the con- 
tinent. Giant business corporations were organized to build up the iron and 
steel, oil, farm equipment, copper, lumber, transportation, and many other basic 
industries on a continental scale. During the last quarter of the century all 
energies were turned to the creation of an industrial society. 

The broad outlines of this industrial transformation will be represented by 
three historical series: 1) a measure of changes in regional populations, 2) a 
measure of changes in the structure of regional production, and 3) a measure 
of accompanying changes in regional income and material well-being. The first 
of the series traces the occupation and settlement of the continental territory, the 
second shows how the populations of the various regions were employed, and the 
third presents the outcome of these changes in terms of total and per capita 
personal income. 



Growth and Shift of Population, 1870-1910 

Between 1870 and 1910 the population of the United States more than 
doubled, rising from less than 40 millions to more than 91 millions. This 
increase was marked by a steady decline in the rate of population growth, 



122 / 



Changing Regional Structure of U. S. Economy, 1870-1910 (1) / 123 

although the annual increase averaged one and a third millions down almost to 
1920. Migration was pronounced; each decennial census down through World 
War I reported that more than 20% of native white Americans had moved from 
the state of their birth. 1 In less than thirty years the vast land area between the 
frontier line of 1860 and the Pacific was overrun by population and the historic 
"frontier" line had ceased to exist. But when this fact was first reported in the 
11th United States Census, the territory beyond the 95th meridian was by no 
means closely settled; large pockets of land were not yet put to productive 
purposes and in some parts good farming land was still available as late as 
1910. Even then many areas in the semi-arid West supported only a thin 
veneer of permanent settlement. 

The relative rates of regional population growth for the period 1870-1910 
are shown in Table 22. The four regions of the greater West (Mountain, South- 
west, Plains, and Far West) were the fastest growing areas of settlement in the 
period before 1890; and with the exception of the Plains, these same regions 
continued to grow at a relatively higher rate than the rest of the nation until 
well into the twentieth century. Meanwhile, the older settled areas of the South- 
east and the Northeast had steadily increased their populations, but at a much 
slower rate than the West; only the Middle Atlantic region here surpassed the 
national rate of population growth before 1910. 



Table 22. Percentage 


increases 


in population by census 


decades, 


by region, 


1870-1910 












Region 


1870-80 


1880-90 


1890-1900 


1900-10 


1870-1910 


United States 


26.0 


25.5 


20.7 


21.0 


131.Q 


New England 


15.0 


17.2 


19.0 


17.2 


87.9 


Middle Atlantic 


19.4 


20.3 


20.9 


23.6 


114.7 


Great Lakes 


22.8 


20.3 


18.6 


14.2 


100.0 


Southeast 


17.8 


17.8 


18.3 


15.6 


89.7 


Plains 


59.7 


45.1 


15.8 


12.5 


201.8 


Southwest 


73.1 


56.6 


51.6 


46.4 


501.4 


Mountain 


151.4 


113.1 


43.1 


53.7 


1,078.8 


Far West 


64.0 


64.5 


27.0 


73.8 


495.6 



Figures in italics are above the national average. 
Source: 17th U. S. Census 1950, see supra, Table 1, note. 



States and regions which experienced the most rapid growth were, almost 
without exception, the least populated parts of the country. Table 23 shows 
population per square mile of land area in 1870, 1890, and 1910 for the regions 

*C. J. Galpin and T. B. Manny, Interstate Migration Among the Native White Population 
as Indicated by Difference between State of Birth and State of Residence, U. S. Department 
of Agriculture, Bureau of Agricultural Economics, Washington: Government Printing 
Office, 1934) , pp. 6-7. 



124 / Regional Economic Development: 1870—1950 



Table 23. Density of population, by region, 1870, 1890, and 1910 

(population per square mile of land area) 



Region 


Area* 


1870 


1890 


1910 




2,974,726 








United States 


sq. mi. 


13.4 


21.2 


30.9 


New England 


( 2.12) 


55.2 


74.4 


103.7 


Middle Atlantic 


( 3.78) 


87.6 


125.8 


188.1 


Great Lakes 


( 8.23) 


37.3 


55.0 


74.5 


Southeast 


(17.95) 


21.7 


30.1 


41.2 


Plains 


(17.17) 


7.6 


17.5 


22.8 


Southwest 


(19.08) 


1.8 


4.8 


10.7 


Mountain 


(17.23) 


0.3 


1.8 


3.9 


Far West 


(14.44) 


1.7 


4.5 


9.9 



* Figures in parentheses show proportion of total U. S. land area 1950. 
Figures in italics are above the national average. 

Source: Statistical Abstract of the United States, 1954, Table 4, p. 10; 17th U. S. Census 
1950, see supra, Table 1, note. 



Table 24. Regional distribution of U. S. population, 1870 - 1910 

(per cent of total) 



Region 


1870 


1880 


1890 


1900 


1910 


United States 


39,818,449 


50,155,783 


62,947,714 


75,994,575 


91,972,266 




100% 


100% 


100% 


100% 


100% 


New England 


8.8 


8.0 


7.5 


7.4 


7.1 


Middle Atlantic 


24.7 


23.4 


22.5 


22.5 


23.0 


Great Lakes 


22.9 


22.3 


21.4 


21.0 


19.8 


Southeast 


29.1 


27.3 


25.5 


25.1 


23.9 


Plains 


9.7 


12.3 


14.2 


13.6 


12.7 


Southwest 


2.5 


3.5 


4.4 


5.5 


6.6 


Mountain 


0.4 


0.9 


1.4 


1.7 


2.2 


Far West 


1.8 


2.3 


3.1 


3.2 


4.7 



Source: 17th U. S. Census 1950, see supra, Table 1, note. 



and for the nation as a whole. Except for the Middle Atlantic region after 1890, 
the regions with population density higher than the national average were those 
whose rate of population growth was below the national average throughout the 
period under review. In absolute terms, of course, the Middle Atlantic, Great 
Lakes, New England, and Southeast regions persistently had the heaviest 
increases in population, though by 1910 their combined share of national popu- 



Changing Regional Structure of U. S. Economy, 1870-1910 (1) / 125 

lation had fallen almost 14%. The effects of these disparate rates of growth on 
the percentage distribution of total population among the regions is shown in 
Table 24. All regions west of the Mississippi increased their shares of national 
population in each successive decade between 1370 and 1910 except the Plains, 
which reached its peak in 1890 and declined thereafter. All regions east of the 
Mississippi experienced declining shares of national population through the 
same period, except the Middle Atlantic region, which held its position after 
1890 and actually increased its share of the national total during the opening 
decades of the present century. 




Absolute upward shift 
' E-X-.'v'| Relative downward shift 



Figure 22. Net Shift in Total Population, 1870-1910. 
State figures represent % of total net shift 



Total net shift as % 



of 



incremental change: 25.4. Total net shift as % of 1870 population: 33.2. 



Figure 22 shows the shift of population among the states during the period 
1870-1910. 2 The heaviest net upward shifts occurred in the Far West, South- 
west, and Plains states, and the heaviest net downward shifts in the Ohio Valley, 
the New York area, the Upper South, and the older settled Plains states, such as 
Missouri and Iowa; Maine also experienced a sizable net downward shift. These 
upward and downward tendencies accord closely with the popular notion of the 
"westward movement." The principal net shifts are summarized in Table 25. 

2 The state figures are based on net shifts, employing the technique described in Chapter 
3; that is, the percentage increase in national population over the 40-year period is applied 
to each state's population in 1870 to obtain what might be called the expected change by 
1910 if population had increased at a uniform rate throughout the states and territories, 
and this expected population for each state is then compared with the actual population 
as recorded by the United States census for 1910. The change in relative position is set 
down as an "upward" or "downward" shift. 



126 / Regional Economic Development: 1870-1950 

A further feature of population changes after 1870 is the marked decline in 
the proportion of total population living in rural areas and the corresponding 
increase in urban population. In every decade between 1870 and 1910 all eight 
regions showed a decline in the rural proportion of their total populations. The 
rate of increase in urban population was positive at every census during the 
40-year period. 

Table 25. Principal net shifts in population among states, 1870 - 1910 





Percentage net 




Percentage net 




upward shift 




downward shift 


Texas 


13.57 


Ohio 


10.50 


Oklahoma 


12.53 


Indiana 


8.93 


Washington 


8.22 


Kentucky 


8.34 


California 


8.19 


Virginia 


8.21 


Minnesota 


8.02 


Tennessee 


7.92 


Nebraska 


6.87 


New York 


7.63 


Kansas 


6.42 


Maine 


5.34 


Colorado 


5.35 


Missouri 


5.16 


N. Dakota 


4.32 


N. Carolina 


4.12 


S. Dakota 


4.21 


Iowa 


4.03 



Source: See Table 1, note. 

Table 26. Urban population as percentage of total population, by region, 
1870 and 1910 



Region 



1870 



1910 



United States 

New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 



25.2 



45.7 



44.4 
44.1 
21.6 

9.5 
18.9 

6.9 
13.9 
31.2 



73.3 
70.2 
52.7 
19.4 
33.2 
22.5 
40.7 
56.0 



Figures in italics are above the national average. 
Source: 17th U. S. Census, 1950, see supra, Table 4, note. 



The proportion of urban to total population in each of the regions in 1870 
and in 1910 is shown in Table 26, and decennial rates of urban increase, 
by regions, in Table 27. Read in conjunction, the two tables show that the 
regions with the highest rates of urbanization are the "newer" and the less 
developed regions. 



Changing Regional Structure of U. S. Economy, 1870-1910 (1) / 127 



Table 27. Decennial rates of increase of urban population, by region, 
1870-1910 



Region 


1870-80 


1880-90 


1890-1900 


1900-10 


United States 


41.0 


56.5 


36.4 


39.3 


New England 


35.7 


13.8 


32.5 


25.3 


Middle Atlantic 


34.7 


39.6 


35.0 


34.6 


Great Lakes 


56.2 


66.0 


41.2 


33.2 


Southeast 


16.2 


63.8 


37.1 


48.3 


Plains 


53.1 


106.2 


27.7 


31.3 


Southwest 


133.5 


135.2 


165.9 


118.3 


Mountain 


351.5 


197.9 


51.6 


68.6 


Far West 


89.1 


94.1 


37.4 


112.1 



Figures in italics are above the national average. 
Source: 17th U. S. Census 1950, see supra, Table 4, note. 



The data reveal a tendency for rates of urbanization to fall off over the period 
as a whole in all regions except the Southeast and the Far West. They also 
reveal a cyclical tendency; decades of relatively rapid urbanization are almost 
always followed by decades of relatively slower urbanization, with the rate of 
over-all urban increase slackening in each successive cycle. 

Table 28 compares the rates of increase in urban population for the eight 
regions over the 1870-1880 and 1900-1910 decades and shows the proportion 
of the nation's total urban population in each region in 1870 and in 1910. 

The net shifts in urban population between 1870 and 1910 are shown for the 
states in Figure 23 and the principal shifts in Table 29. Some parts of the more 



Table 28. Rank order of regions by rate of increase in urban population 
1870 - 1880 and 1900 - 1910, and regional proportions of U. S. urban 
population, 1870 and 1910 



1870-80 1870 1900-10 1910 

Region % Rate Proportion % Rate Proportion 

(1) Mountain 

(2) Southwest 

(3) Far West 

(4) Great Lakes 

(5) Plains 

(6) New England 

(7) Middle Atlantic 

(8) Southeast 

Rank order in parentheses. 
Source: 17th U. S. Census 1950, see supra, Table 4, note. 



351.5 


0.2 (8) 


68.6 (3) 


2.0 (8) 


133.5 


0.7 (7) 


118.3 (1) 


3.3 (7) 


89.1 


2.2 (6) 


112.1 (2) 


5.7 (6) 


56.2 


19.7 (2) 


33.2 (6) 


22.9 (2) 


53.1 


7.3 (5) 


31.3 (7) 


9.2 (5) 


35.7 


15.5 (3) 


25.3 (8) 


11.4 (3) 


34.7 


43.4 (1) 


34.6 (5) 


35.3 (1) 


16.2 


11.0 (4) 


48.3 (4) 


10.2 (4) 



128 / Regional Economic Development: 1870-1950 




Absolute upword shift 
I'--'v/:[ Relative downward shift 



Figure 23. Net Shift in Urban Population, 1870-1910. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 22.0. Total net shift as % of 1870 urban population: 70.1. 



recently settled areas of the country — the Far West states and Texas — are seen 
to have had sizable upward shifts in urban population over these years. The 
greatest relative increase in any one state, however, was in Illinois, chiefly in 
the Chicago metropolitan area. Minnesota and Michigan also show considerable 
increases which are connected, as in Illinois, with the growth of their dominant 
metropolitan centers. The greatest downward shifts were in states in the highly 

Table 29. Principal net shifts in urban population among states, 1870 - 1910 





Percentage 


net 




Percentage net 




upward shift 




downward shift 


Illinois 


13.97 




New York 


28.31 


Texas 


9.73 




Massachusetts 


15.35 


Washington 


8.62 




Pennsylvania 


12.41 


California 


8.46 




Maryland and D. C. 


10.74 


Minnesota 


7.88 




Louisiana 


6.37 


Colorado 


5.44 




Missouri 


5.79 


Michigan 


4.69 




Kentucky 


5.09 


Oklahoma 


4.54 




Maine 


4.13 


New Jersey 


3.97 




Virginia 


2.88 


Kansas 


3.91 




Ohio 


2.81 



Source: 17th U. S. Census 1950, see supra, Table 4, note. 



Changing Regional Structure of U. S. Economy, 1870-1910 (1) / 129 

urbanized New England and Middle Atlantic regions. In the Great Lakes region, 
however, only Ohio shows any noteworthy relative decline in its rate of urban 
growth. 

In absolute terms, however, urbanization was most marked in states which 
developed concentrations of manufactures and related service activities. From 
the standpoint of social change, these centripetal currents of population were 
more decisive than the centrifugal currents of "the westward movement." By 
1900 only six in every ten people were classified as "rural," and before 1920 
less than half the population of 105,000,000 could be credited to the countryside. 
The United States had become an urban-industrial society; every other Ameri- 
can, native-born and immigrant, lived out his life in an urban environment. 



12, / Changing regional structure of the 
U. S. economy, 1870-1910: 

(2) SHIFTS IN PRODUCTIVE ACTIVITIES 



The continuing growth and redistribution of population within the United States 
after 1870 was accompanied by changes in the size, occupational structure, and 
geographical distribution of the labor force. 1 There is perhaps no better single 
index of changes in the growth and distribution of productive activities, of 
society's economic base, then the changing pattern of employment through which 
a population earns its living. 

The major categories used for this structural analysis have already been 
suggested in our description of the economy in 1870; they are three in number, 
having reference to successive stages in the production process: 

1) Resource or Primary Activities, which are concerned with the extraction 
of raw materials. They embrace such varied occupations as agriculture, forestry 
and logging, mining, and fishing. 

2) Processing or Secondary Activities, which have to do with the trans- 
formation of raw materials into finished or semi-finished goods — in short, with 
manufactures and associated mechanical trades. 

3) Servicing or Tertiary Activities, which are essentially the "all other" types 
of employment engaged in by the residual component of the labor force, i.e., by 
workers not engaged in either the extraction of materials or the manufacture of 
products. 

Together, these three categories — primary, secondary, and tertiary activities 
— comprise the total labor force available for production in the nation or any of 
its regions. 

The growth of the national labor force and its principal components in the 
United States between 1870 and 1910 is shown in Table 30, and its changing 
regional distribution over these years in Table 31. In size and distribution, as 
might be expected, the labor force at any time is roughly proportionate to the 

a For present purposes the term "labor force" includes all active workers, ten years old 
and over, regularly [since 1940, currently at the time of the census] employed in a given 
occupation. 

130 / 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 131 

total population. 2 Thus in 1870, when the westward movement of population was 
in full swing, the four older regions east of the Mississippi had about 85% of 
the nation's total population and almost exactly the same percentage of its total 
labor force. By 1910 their share in the population had dropped to a little over 
73% and their share of the labor force to slightly less than 75% of the national 
total. Most of this decline in the share of the labor force — approximately two- 
thirds of it — took place before the symbolic closing of "the frontier" in 1890. 
The shift of some 10% of the greatly expanded labor force west of the Mis- 
sissippi over the 40-year interval is a cumulative measure of the westward 
movement in terms of job opportunities. 



Table 30. U. S. labor force, distribution by major industry group, 1870, 1890, 
and 1910 



Year 


Labor force 


% 
Resource 


% 
Manufactures 


% 
Services 


1870 
1890 
1910 


12,505,923 
22,735,661 
38,167,336 


53.45 
43.39 
35.62 


21.14 
24.30 
27.92 


25.41 
32.31 
36.46 



Source: Appendix Tables A1-A7. 



Table 31. Regional distribution of U. S. labor force, 1870, 1890, and 1910 



Region 


1870 


1890 


1910 


United States 


12,505,923 


22,735,661 


38,167,336 




100% 


100% 


100% 


New England 


10.39 


8.82 


7.64 


Middle Atlantic 


25.23 


24.30 


23.56 


Great Lakes 


21.90 


20.61 


19.02 


Southeast 


28.04 


24.09 


24.51 


Plains 


9.24 


13.14 


11.66 


Southwest 


2.18 


3.51 


6.20 


Mountain 


.57 


1.75 


2.24 


Far West 


2.45 


3.78 


5.19 



Source: Appendix Table A-l. 



Within the emergent greater West, the growth of the labor force before 1890 
was fastest in the Mountain states but largest in the Plains; the Southwest and 
Far West grew at roughly the same rate and to the same extent (see Table 31). 

: The proportion of the total population in the labor force, however, was somewhat higher 
at the end of the period under review than at the beginning — about two-fifths in 1910 as 
compared to about one-third in 1870. 



132 / Regional Economic Development: 1870-1950 

Between 1890 and 1910, however, the Plains region passed its peak share, and 
by 1900 was already in relative decline; but the other regions continued to 
increase their shares of the growing national labor force. The Southwest grew 
somewhat faster than the Mountain or the Far West regions during these years, 
and by 1910 had overtaken the Far West as the second largest western region 
after the Plains. 

The westward movement, again, had a differential impact on the relative labor 
force shares of the four older regions in the East. Between 1870 and 1890, the 
shares of New England and the Southeast both declined by 14% ; in contrast, 
the decline in the Great Lakes region was only 5% and that in the Middle 
Atlantic region about 4%. Between 1890 and 1910, the Southeast actually 
increased its share of the total labor force slightly, but the shares of the other 
three regions continued in decline. 

The differential effects of the redistribution of population and labor force 
among the regions over the 1870-1910 period as a whole are summarized in 
Table 32. 



Table 32. Percentage change in regional shares of U. S. population and labor 
force, 1870 - 1910 



Region Population Labor force 

New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 

Source: 17th U. S. Census, 1950, see supra, Table 1, note. Appendix Table A-l. 



Resource Activities in the Regions, 1870-1910 

There were major shifts in the location of primary resource activities within 
the national economy between 1870 and 1910. 3 Most striking perhaps was the 
Northeast's relative decline after about 1890 (Table 33). Its share in the 
national labor force engaged in resource extraction fell from 44% of the total 
in 1870 to 36% in 1890 and to only 27% in 1910. All four regions of the 
greater West had larger shares of the nation's resource activity in 1910 than in 
1870. However, in the Plains region, though some net gain took place over the 

3 Moses Abramovitz, Resource and Output Trends in the United States since 1870, 
Occasional Paper 52 (New York: National Bureau of Economic Research, 1956), pp. 5-23. 



- 18.72 


- 26.47 


- 7.04 


- 6.62 


- 13.44 


- 13.15 


- 17.85 


- 12.59 


+ 30.55 


+ 26.19 


+ 160.63 


+ 184.40 


+ 411.63 


+ 292.98 


+ 158.33 


+ 111.84 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 133 

period as a whole, the share of the total primary labor force was smaller in 1910 
than in 1890. The Far West's relative increase was also surprisingly small after 
1890. 

Table 34 presents a regional breakdown of labor force in resource activity, 
classified in the four principal sectors: agriculture, mineral extraction, forestry 
and logging, and fisheries. In agriculture the Northeast was in continuous 
decline over the entire 1870-1910 period; this decline was especially marked 
in New England and the Middle Atlantic states. The combined share of the 
Northeastern regions in the nation's agricultural labor force fell from 43% of 
the total in 1870 to 34% in 1890 and only 24% in 1910. In contrast, the South- 
east's share rose in every decade but one (1890) ; by 1910 more than 44% of 
the nation's total agricultural labor force was employed in the Southeast. The 
expansion here, however, was more than offset by the decline in the Northeast. 
As a result, the greater West's share of farm labor shows an increase for every 
decade between 1870 and 1910. 4 



Table 33. Regional distribution of all U. S. resource industries labor force, 
1870, 1890, and 1910 



Region 1870 1890 1910 

United States 



New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 



6,684,219 


9,864,596 


13,596,815 


100% 


100% 


100% 


5.43 


3.65 


2.29 


14.85 


12.33 


9.68 


23.44 


20.15 


15.05 


39.71 


37.68 


42.34 


10.80 


16.12 


14.07 


3.02 


5.29 


10.41 


.64 


1.63 


2.45 


2.10 


3.15 


3.71 



Source: Appendix Table A2-A5. 



The bulk of the westward shift in agriculture occurred before 1890, as is 
evidenced by the fact that the agricultural labor force in the greater West 
already comprised 26% of the national total in 1890 compared with less than 
16% in 1870 (and some 30% in 1910). Moreover, all four western regions 
greatly increased their shares of the agricultural labor force between 1870 and 
1890; the impact was felt heavily in the Plains during the seventies and in all 
the regions during the eighties. The share of the Plains peaked in 1890, as did 
that of the Far West; but while the Plains continued in decline after 1900, the 
Far West had more than recovered its 1890 position by 1910. The Mountain 

"This period, of course, was one in which there were generally increasing amounts of 
labor used in agriculture. 



134 / Regional Economic Development: 1870-1950 

region increased its share of farm workers over every census decade during this 
period. But the most striking growth was that in the Southwest, especially in 
Texas and Oklahoma around the turn of the century. By 1910 the proportion of 
the agricultural labor force in the Southwest exceeded that of the New England 
and Middle Atlantic regions combined. 

In the second largest resource sector, mining employment, there was con- 
siderably less regularity in the changing patterns of regional distribution 
between 1870 and 1910. Mining developed more rapidly in relative terms than 
agriculture, but in some of the regions also declined more spectacularly as fixed 
mineral deposits were used up. 

Mining was the most western of all economic activities in 1870. At that time 
the four regions west of the Mississippi had 43% of the nation's mineral labor 
force as compared with only about 15% of the total population, about the same 
percentage of the total labor force, and approximately 16% of the agricultural 
labor force of the country. About 40% of all mining employment was concen- 
trated in the Mountain and Far West regions. But after 1870, in contrast to the 
general westward movement, the trend of mining was eastward. During the 
eighties the Far West's share in mining employment declined by one-half, and 
by 1890 the greater West had only 30% of the nation's mining workers. By 
1910 the proportion had fallen to 26% ; almost 74% of the total mining force 
was now employed in the regions east of the Mississippi. This change suggests 
the rising importance of coal mining during the last quarter of the century. 

The movement of labor in forestry and logging activity after 1870 resembles 
that both of general population and labor force in its shift to the West. Between 
1880 and 1890, however, there was an even more pronounced shift into the 
Southeast, more nearly resembling that in mining activity. Much of the South- 
ern expansion seems to have stemmed directly from the destruction of the pine 
forests in the northern Great Lakes states. By 1900, in fact, the entire North- 
eastern share of logging activity had been greatly reduced. In 1870 the North- 
east accounted for almost two-thirds of all forest employment. By 1910 its 



Table 34. Regional 


distribution 


of each 


resource 


industry's 


labor force, 


1870 - 1910 




Agriculture 








Region 


1870 


1880 


1890 


1900 


1910 


United States 


6,437,372 


8,590,280 


9,235,290 


11,288,027 


12,389,840 




100% 


100% 


100% 


100% 


100% 


New England 


5.32 


3.93 


3.56 


2.78 


2.27 


Middle Atlantic 


14.26 


11.70 


11.01 


8.97 


7.54 


Great Lakes 


23.71 


21.36 


20.09 


17.83 


14.74 


Southeast 


40.93 


41.51 


39.32 


41.73 


44.50 


Plains 


11.07 


14.35 


16.68 


16.32 


14.67 


Southwest 


3.12 


4.93 


5.52 


8.38 


11.01 


Mountain 


.37 


.52 


1.14 


1.40 


2.06 


Far West 


1.22 


1.70 


2.68 


2.59 


3.21 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 135 



Mineral Extraction 



Region 


1870 


1880 


1890 


1900 


1910 


United States 


186,616 


297,784 


447,001 


694,352 


965,169 




100% 


100% 


100% 


100% 


100% 


New England 


2.85 


2.44 


2.11 


1.17 


.94 


Middle Atlantic 


33.36 


34.61 


35.91 


34.49 


36.68 


Great Lakes 


16.54 


13.01 


21.43 


19.56 


19.61 


Southeast 


4.10 


4.84 


10.07 


13.28 


16.73 


Plains 


3.08 


6.44 


8.64 


8.97 


8.40 


Southwest 


.45 


2.53 


2.03 


3.37 


4.33 


Mountain 


10.05 


16.65 


11.28 


11.12 


7.49 


Far West 


29.57 


19.48 


8.53 


8.04 


5.82 



Forestry and Logging 



Region 


1870 


1880 


1890 


1900 


1910 


United States 


32,360 


55,931 


122,143 


140,599 


173,531 




100% 


100% 


100% 


100% 


100% 


New England 


10.94 


6.61 


5.63 


4.97 


5.23 


Middle Atlantic 


20.61 


14.37 


17.01 


12.41 


7.77 


Great Lakes 


23.66 


33.93 


26.14 


17.80 


13.00 


Southeast 


16.84 


14.85 


21.31 


33.04 


35.13 


Plains 


9.45 


9.21 


7.96 


8.83 


7.32 


Southwest 


1.31 


2.83 


1.96 


3.87 


4.93 


Mountain 


1.83 


3.24 


3.82 


2.71 


3.31 


Far West 


15.36 


14.96 


16.17 


16.37 


23.31 



Fisht 



Region 


1870 


1880 


1890 


1900 


1910 


United States 


27,871 


41,352 


60,162 


68,940 


68,275 




100% 


100% 


100% 


100% 


100% 


New England 


40.32 


30.78 


25.32 


18.07 


18.23 


Middle Atlantic 


21.53 


25.96 


31.25 


27.37 


21.21 


Great Lakes 


7.84 


6.64 


7.37 


8.64 


11.03 


Southeast 


24.38 


18.53 


25.51 


30.19 


31.71 


Plains 


1.15 


1.16 


1.97 


2.95 


2.30 


Southwest 


.21 


.35 


.74 


1.10 


1.71 


Mountain 


.04 


.07 


.11 


.08 


.22 


Far West 


4.53 


16.51 


7.73 


11.60 


13.59 



Source: Appendix Tables A2-A5. 






136 / Regional Economic Development: 1870-1950 

share was only about one-fourth, while the Southeast had more than one-third 
and the Far West almost one-fourth of the national labor force in this important 
resource sector. 

The data on fisheries employment show a pattern of regional redistribution 
quite different from that seen in the other primary industries. In 1870 the 
New England and Middle Atlantic regions together accounted for almost 62% 
of the nation's fishing activity. In 1890 these two regions still employed more 
than 56% of the total fisheries labor force. New England's share in fishing 
employment declined during these two decades, but this decline was partly offset 
by a rise in the Middle Atlantic fisheries, which reached their peak in 1890. The 
Southern fisheries and those of the Great Lakes region barely held their own 
between 1870 and 1890; and in the Far West, rapid growth during the seventies 
was followed by an almost equally rapid relative decline during the eighties. 
After 1890, however, the shares of the Southeast and Far West in the fisheries 
labor force increased considerably, while those of the New England and Middle 
Atlantic regions continued in gradual relative decline. Elsewhere, only the Great 
Lakes region showed much incremental activity. Its share rose from around 
7% before 1890 to more then 11% by 1910. 

Table 35 summarizes changes between 1870 and 1910 in the share of the 
Northeast, the Southeast, and the West in each of the four major categories of 
resource employment. The Northeast declined relatively over these years in all 
categories except mineral extraction. The Southeast gained relatively in all and 
most notably in mineral extraction, which increased fourfold, and in forest 
employment, which more than doubled. The West gained relatively in all cate- 



Table 35. Percentage distribution of each resource industry's employment, by 
great region, 1870 and 1910 



Region 


Industry 


Year 




1870 


1910 


Northeast 


Agriculture 


43.29 


24.55 




Mineral extraction 


52.75 


57.23 




Forestry and logging 


55.21 


26.00 




Fisheries 


69.69 


50.47 


Southeast 


Agriculture 


40.93 


44.50 




Mineral extraction 


4.10 


16.73 




Forestry and logging 


16.84 


35.13 




Fisheries 


24.38 


31.71 


West 


Agriculture 


15.78 


30.95 




Mineral extraction 


43.15 


26.04 




Forestry and logging 


27.95 


38.87 




Fisheries 


5.93 


17.82 



Source : Appendix Tables A2-A5. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 137 

gories except mineral extraction, most notably in fishing and agriculture, which 
trebled and doubled, respectively; mining activities declined relatively by more 
than one-third. 



Value of Resources Extracted, 1870-1910 

When the data on value of resources extracted are examined, the regional 
shares of the national totals assume somewhat different proportions from those 
shown by the data on labor force employed in resource activities. 5 The most 
striking change again is the continuous decline in the shares of the Northeast- 
ern regions (Table 36). In 1870 these three regions contributed more than 58% 



Table 36. Regional distribution of value of primary U. S. resources extracted, 
1870, 1890, and 1910 



Region 


1870 


1890 


1910 


United States 


$2,745,463* 


$3,280,974* 


$10,512,395* 




100% 


100% 


100% 


New England 


6.8 


4,7 


3.2 


Middle Atlantic 


23.7 


17.3 


11.6 


Great Lakes 


27.7 


24.2 


20.6 


Southeast 


25.2 


21.7 


22.4 


Plains 


11.1 


18.9 


24.5 


Southwest 


1.9 


4.1 


7.6 


Mountain 


.5 


3.9 


4.1 


Far West 


3.1 


5.2 


6.0 



*(000's of current dollars). 
Source: Appendix Table B-l. 



of the total primary resource value, but by 1910 their combined share had fallen 
to little more than 35%. The Southeast also suffered a relative decline although 
the low point was reached before 1890 and its share recovered somewhat there- 
after. The regions of the greater West, on the other hand, improved their rela- 
tive shares between 1870 and 1890 and again down to 1910. The rate of relative 
gain, however, was greater in the earlier period (it doubled) than in the later 
period (less than one-third). A comparison of these value components with 

5 The figure for the value of extracted resources is a rather crude aggregate of the follow- 
ing items culled from official sources: the gross value of farm production; the gross value 
of mineral production; the gross value of fish at point of sale by fishermen (excluding the 
whale fishery) ; and the average value of hardwood and softwood cut for timber plus the 
value of domestic pulpwood. These figures are not strictly comparable from decade to 
decade (quite apart from changes in the value of money), but they do provide a fairly 
accurate geographical cross-section of current values produced at any one time. 



138 / Regional Economic Development: 1870-1950 



Table 37. Distribution of. labor force in resource industries, and value of 
extracted resources, by great region, 1870, 1890, and 1910 



Region 




1870 






1890 




1910 


United States 


L.F. 

100% 




Value 
100% 


L.F. 

100% 




Value 
100% 


L.F. Value 
100% 100% 


Northeast 
Southeast 
West 


43.7 
39.7 
16.6 




58.2 
25.2 
16.6 


36.1 
37.7 
26.2 




46.2 
21.7 
32.1 


27.0 35.4 
42.3 22.4 
30.7 42.2 



Source: Appendix Tables A2-A5 and B-l. 



shares of the primary labor force reveals some striking discrepancies, most 
notably in the Southeast and the greater West (Table 37) . 

The Southeast's rising share of the primary labor force after 1890 was in no 
way matched by a corresponding increase in its share of the value of the nation's 
extracted resources. From the standpoint of the value of primary resources, 
this region was relatively worse off in 1910 than at the height of Reconstruction. 
The West, on the other hand, experienced a large increase in its share of the 
primary labor force (more than 80% ) but an even greater increase in the value 
of its extracted resources (a rise of more than one and a half times). The 
Northeast managed to hold its declining proportions of both primary labor force 
and value of resources in approximate balance. 

These divergencies underline the fact that, for a variety of reasons, labor 
productivity differs markedly among the several resource sectors and that within 
each sector it varies from region to region. Thus, since the eighties, the strength 
of resource extraction in the West has lain in the proportion of its high value 



Table 38. Regional distribution of value of resources extracted by major 
resource industry 1870, 1890 and 1910 

Agriculture 



Region 


1870 


1890 


1910 


United States 


$2,447,541* 


$2,460,003* 


$8,494,231* 




100% 


100% 


100% 


New England 


6.29 


4.33 


2.89 


Middle Atlantic 


21.41 


14.03 


9.13 


Great Lakes 


28.26 


23.07 


21.63 


Southeast 


27.44 


25.51 


21.83 


Plains 


11.75 


22.18 


28.33 


Southwest 


2.10 


4.67 


8.04 


Mountain 


.27 


1.23 


3.17 


Far West 


2.48 


4.98 


4.98 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 139 



Mineral extraction 



Region 


1870 


1890 


1910 


United States 


$152,596* 


$582,133* 


$1,238,415* 




100% 


100% 


100% 


New England 


3.03 


3.86 


1.40 


Middle Atlantic 


56.73 


32.46 


30.45 


Great Lakes 


15.44 


23.11 


19.18 


Southeast 


3.27 


7.30 


12.90 


Plains 


3.13 


8.19 


10.52 


Southwest 


.24 


2.61 


6.15 


Mountain 


5.07 


16.62 


11.46 


Far West 


13.09 


5.85 


7.94 


Forestry and logging 


Region 


1870 


1890 


1910 


United States 


$134,229* 


$199,937* 


$725,737* 




100% 


100% 


100% 


New England 


15.19 


6.56 


7.78 


Middle Atlantic 


28.49 


9.63 


7.36 


Great Lakes 


32.41 


45.51 


12.88 


Southeast 


11.41 


16.95 


45.48 


Plains 


8.51 


12.60 


5.18 


Southwest 


.52 


2.40 


4.95 


Mountain 


.55 


.58 


2.23 


Far West 


2.92 


5.77 


14.14 


Fisheries 


Region 


1870 


1890 


1910 


United States 


$11,097* 


$38,901* 


$54,012* 




100% 


100% 


100% 


New England 


72.90 


27.16 


28.03 


Middle Atlantic 


7.63 


38.40 


22.26 


Great Lakes 


10.67 


5.17 


9.33 


Southeast 


4.28 


19.43 


25.57 


Plains 


.04 


1.07 


1.32 


Southwest 


— 


.82 


.83 


Mountain 


.02 


.20 


— 


Far West 


4.46 


7.75 


12.66 



*(000's of current dollars). 
Source : Appendix Tables B2-B5. 



resource activities rather than in the absolute size of its primary labor force. 
Table 38 shows the value of the four main categories of resources extracted in 
the regions between 1870 and 1910. The relative decline of the Northeast is 



140 / Regional Economic Development: 1870-1950 

apparent in all categories, although output of mineral wealth in the Middle 
Atlantic region and the strength of the Great Lakes region in agriculture, 
minerals, and forest activities (before 1900) tends to cushion that decline. The 
relative decline in the Southeast's agricultural values in contrast to agricultural 
labor force emphasizes the labor-intensive nature of the region's agriculture. 
Labor force components in mining and fishing in the Southeast also make up a 
larger proportion of the national totals in these sectors than do the value com- 
ponents that result from labor; the disparity, however, is much less marked than 
in agriculture. The most favorable aspect of the Southeast's resource economy 
before 1910 is the strong rise of its forest industries: the share of the forest 
labor force increased by almost two-thirds between 1890 and 1910, but the 
relative value of the raw forest product increased by more than 168%. 

Among the Western regions at this time, the most spectacular achievement in 
the Plains was the rise of agriculture in the face of many natural obstacles; the 
share of the farm labor force between 1870 and 1910 rose by a little more than 
one-third, but the relative share of farm values more than doubled. Certainly 
no other Western region contributed as much to the nation's agriculture in these 
years. On the other hand, all Western regions made important contributions to 
the exploitation of mineral wealth. By 1910 the Mountain and Plains regions had 
achieved the larger shares, although the former was already in relative decline; 
but rapid growth in the Southwest (notably Texas) after 1890 and the revival 
of mineral interest in the Far West (notably California) marked the appearance 
of new and valuable oil fields in the nation's resource economy. The riches of 
the Far West (especially the Pacific Northwest) in forest resources also began 
to make their mark around the turn of the century. This sector was still very 
labor-intensive as late as 1910, but the share of raw forest values was beginning 
to rise much more rapidly than the share of the forest labor force. Structural 
differences in the primary labor force plus differences in labor productivity 
and product prices help account for the varying contributions of the regions to 
the nation's resource economy. 

Taking the nation's resource economy as a whole during this period, what 



Table 39. Percentage distribution of value of all resources extracted, by major 
resource industry, 1870 and 1910 



Resource industry 



1870 



1910 



All resource industries 



Agriculture 
Mineral extraction 
Forestry and logging 
Fisheries 



$2798,637* 


$14,827,073* 


100% 


100% 


89.15 


80.80 


5.56 


11.78 


4.89 


6.90 


.40 


.52 



*(000's of 1929 dollars). 
Source: Appendix Tables B2-B5. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 141 

changes in the four resource sectors can be linked to the development of re- 
sources in the three great regions? Between 1870 and 1910 the real value of 
resources extracted increased fivefold; the value (in terms of 1929 dollars) was 
about $2,798,637,000 in 1870 and over $14,827,073,000 by 1910. The con- 
tribution of agriculture declined somewhat during this period as a whole, while 
that of forest activities and especially of mineral activities increased consider- 
ably; fisheries made only a slight contribution at all times (Table 39). Actually, 
the share of agriculture had increased after the recovery from the depression of 
the early nineties and, pari passu, the share of mining had declined after 1890 
in spite of the early rise of the petroleum industry. Only forestry, in fact, 
appears to have improved its relative position throughout the entire period. 

These developments marked important geographical shifts in the nation's 
resource economy. Thus, apart from the Great Lakes states, the recovery of 
agriculture after the depression of the nineties was confined essentially to the 
greater West, especially to the Plains and Southwest. Neither the Northeast nor 
the Southeast raised its relative agricultural contribution after 1890; in all the 
territory east of the Mississippi, only the Great Lakes region held its own in the 



Table 40. Percentage distribution of value of all resources extracted, by maj 
resource industry and great region, 1870, 1890, and 1910 



or 



Resource 












industry 




Northeast 


Southeast 


West 


United States 


Agriculture 












1870 




49.89 


24.46 


14.80 


89.15 


1890 




31.06 


19.13 


24.79 


74.98 


1910 




27.19 


17.64 


35.97 


80.80 


Mineral extraction 










1870 




4.18 


.18 


1.20 


5.56 


1890 




10.55 


1.29 


5.90 


17.74 


1910 




6.01 


1.52 


4.25 


11.78 


Forestry and 


logging 










1870 




3.72 


.56 


.61 


4.89 


1890 




3.76 


1.03 


1.30 


6.09 


1910 




1.93 


3.14 


1.83 


6.90 


Fisheries 












1870 




.38 


.01 


.02 


.40 


1890 




.84 


.23 


.12 


1.19 


1910 




.31 


.13 


.08 


.52 



Source: Appendix Tables B2-B5. 



142 / Regional Economic Development: 1870-1950 

agricultural sector. In minerals each of the three great regions of the country 
contributed to the upsurge in total value between 1870 and 1890, although in 
absolute terms the Southeast's share was modest in comparison with that of the 
Northeast or the West. But thereafter the contribution from Northeastern min- 
erals fell off sharply, and even the West experienced a moderate, although tem- 
porary, decrease, whereas the Southeast continued its less spectacular growth 
throughout the entire 40-year period. In the forest sector, the Northeast, the 
West, and the Southeast all increased their relative shares of total resources 
value between 1870 and 1890 (the strength of the Northeast here, as in agricul- 
ture, lay exclusively in the Great Lakes region). After 1890, however, the 
Northeast went into relative decline; the West continued to grow, and the South- 
east by 1910 had more than doubled its relative share (Table 40). 

Attention to the multi-state regions alone conceals some significant changes in 
resource activities among the states within the regions. We shall now examine 
these changes to determine more precisely where, when, and in what magnitude 
the pronounced geographical shifts occurred. 



AGRICULTURE 

The period 1870-1910 was generally one of great instability in agricultural 
markets; but in spite of a secular decline in most farm prices down until the 
late nineties, agricultural settlement continued to spread across the entire 




Figure 24. Net Shift in Agricultural Labor Force, 1870-1910. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 47.7. Total net shift as % of 1870 agricultural labor 
force: 44.1. Source: Appendix Table C. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 143 

country. Figure 24 and Table 41 summarize the principal net upward and 
downward shifts in the agricultural labor force of the various states over these 
four decades. More than a third of the total net upward shift of farm workers 
during these years took place in the Southwest (principally in Texas and, after 
1890, also in Oklahoma), and a further 22% in the wheat belt of the western 



Table 41. Principal net shifts in agricultural labor force among states, by 
decade, 1870-1910 



1870-1880 


1880-1890 


1890-1900 


1900-1910 








Net upward shift 








Tex. 


23.45 


N. & S. 
Daks. 


16.64 


Okla. 


26.01 


Tex. 


15.82 


Kans. 


17.40 


Nebr. 


15.76 


Tex. 


19.64 


Okla. 


12.65 


Ark. 


10.92 


Minn. 


9.77 


Ala. 


9.36 


Ga. 


10.49 


Nebr. 


9.39 


Calif. 


9.10 


Miss. 


8.61 


Miss. 


10.16 


Minn. 


5.47 


Tex. 


6.54 


Ark. 


5.36 


Ark. 


6.59 


La. 


5.00 


Wash. 


5.91 


Ga. 


4.02 


Ala. 


6.34 


S. C. 


4.65 


Kans. 


5.34 


Minn. 


3.66 


N. C. 


5.84 


Iowa 


4.37 


Ark. 


5.30 


Ky. 


3.38 


S. C. 


4.89 


N. & S. 




Colo. 


5.04 


N. C. 


3.31 


N. Dak. 


4.53 


Daks. 


4.24 














Calif. 


2.70 


Wis. 


3.38 


Tenn. 


3.16 


Calif. 


3.34 


Share of 10 














leading 
















states in 
















total net 
















upward 
















shift 


87.59 




82.78 




86.51 




80.65 








Net downward shift 








N. Y. 


19.05 


Ga. 


14.14 


N. Y. 


17.59 


111. 


11.59 


Ohio 


18.21 


111. 


10.43 


Ohio 


12.21 


Mo. 


9.41 


111. 


8.93 


Ala. 


10.15 


Pa. 


9.90 


Ohio 


9.13 


Va. 


7.54 


Ind. 


9.39 


111. 


9.78 


Iowa 


8.58 


Tenn. 


6.67 


Ohio. 


8.57 


Ind. 


7.46 


N. Y. 


7.83 


Pa. 


5.99 


Va. 


8.28 


Mich. 


5.05 


Ind. 


7.69 


Mass. 


5.36 


Ky. 


6.90 


Kans. 


4.84 


Pa. 


7.24 


N. J. 


4.35 


N. Y. 


5.60 


Nebr. 


3.39 


Mich. 


5.52 


Maine 


3.85 


Pa. 


4.95 


Maine 


3.03 


Kans. 


4.41 


Ky. 


3.37 


N. C. 


4.50 


Mass. 


3.03 


Ky. 


3.21 


Share of 10 














leading 
















states in 
















total net 
















downward 
















shift 


83.32 




82.91 




76.28 




74.61 



The figures in italics are absolute downward shifts. 
Source: Appendix Table A-2. 



144 / Regional Economic Development: 1870-1950 

Plains states and Minnesota (mostly before 1890). The spread of cotton-grow- 
ing also brought sizable net upward shifts into certain southeastern states, 
notably Arkansas, Alabama, and Mississippi. 

The absolute and relative declines in agricultural employment were heaviest 
in the more urban-industrial and older agricultural states. The New England, 
Middle Atlantic, and Great Lakes regions lost relative to other parts of the 
country at each decennial census; and after 1890 New England's decline became 
absolute, as did that of several Middle Atlantic and even Great Lakes states after 
1900 (New York, Pennsylvania, Illinois, Ohio, Indiana, and Michigan). At 
various times many Plains and Southeastern states also contributed to the total 
net downward shift, while at others they experienced growth; relative declines 
in the Plains were most marked after 1890, those in the Southeast during the 
eighties. Over the period as a whole, however, Missouri and Iowa were the 
only states in the Plains region and Virginia, Kentucky, and Tennessee the only 
states in the Southeast that showed a net downward movement. 

Agricultural fortunes in both the Plains and the Southeast were adversely 
affected during the last quarter of the nineteenth century by severe price fluctua- 
tions in the national wheat and cotton markets and by prolonged periods of 
drought. Urban industrialization, moreover, did not proceed fast enough in 
these regions to absorb all the labor displaced by mechanization in certain types 
of farming. A further factor complicating the economic position of farmers in 
many parts of the country before 1900 was the fact that, in spite of increases 
in population, demand for farm products grew less rapidly than supply. In 
1870 people spent on an average about one-third of their current incomes on 
agricultural products; by 1890 the proportion had fallen to little more than one- 
fifth, and it was to fall still further before the upturn about 1897. Only a slack- 
ening in the growth of farm population, amounting almost to rural depopulation 
in some parts of New England, and the concurrent decline in cultivated acreage 
and in the growth of farm output after about 1897, together with the regional 
growth of urban population, permitted some gradual amelioration of the staple 
farmer's economic position before World War I. 6 



MINERAL EXTRACTION 

The exploitation of mineral resources has been a vital feature of economic 
development in the United States. After 1870 the nation's draft on coal, iron, 
copper, lead, zinc, and other ores increased enormously. Nature had endowed 
the different regions of the country (except New England perhaps) with an 
immense variety of minerals, most of which were sooner or later adapted to meet 

6 T. W. Schultz, Agriculture in an Unstable Economy (New York: McGraw-Hill, 1945), 
p. 115; H. U. Faulkner, "The Decline of Laissez-Faire, 1897-1917," in The Economic History 
of the United States (New York: Rinehart and Co., 1951), Vol. 7, pp. 315-65. Before 1897, 
however, the position of farmers in the dairy and corn belts of the Great Lakes region 
was only less critical than that of the wheat and cotton producers in the Plains and 
Southeast. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 145 

this growing need of industry for materials and fuels. Given adequate trans- 
portation service, there seemed no limit to the potential supply of domestic 
minerals. Fortunately, the older eastern part of the country had a superb system 
of natural and artificial inland waterways and, by 1870, had the makings of a 
first-class railroad network. Precious metals apart, there was no pressing need 
to develop the mineral resources of the West before railroad transportation had 
become available in the closing decades of the century. Between 1870 and 1910 
the share of minerals in the real value of all extracted resources in the United 
States more than doubled. 

Mining employment rose steadily throughout these four decades, princi- 
pally as a result of greatly increased coal-mining activity. The importance of 
coal mining in the resource economy of the late nineteenth century is evidenced 
by the fact that more than one-third of the total net upward shift in mining 
employment took place in Pennsylvania and West Virginia (see Figure 25). A 



Figure 25. Net Shift in Mining Labor Force, 1870-1910. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 47.8. Total net shift as % of 1870 mining labor force: 
199.5. Source: Appendix Table C. 



further 10% was registered in the coal-producing states of the Great Lakes 
region; Illinois was especially strong during the 1890's although initial develop- 
ments both here and in Indiana can be seen to go back as far as the seventies. 
In the Southeast, the coal and iron district of Alabama showed nearly 8% of 
the total net upward shifts, mostly between 1880 and 1900, and Tennessee about 
4%. Altogether, some 50% or more of the nation's total net upward shift in 
mining employment during the 1870-1910 period was connected with coal 



146 / Regional Economic Development: 1870-1950 

developments, and the half-dozen states concerned all lay east of the Mississippi. 
Nearly two-thirds of the net downward shifts in mining employment over these 
years occurred in the precious metals producing states of the Far West, in spite 
of some revival of activities in Nevada after 1900 and in Washington a decade 
earlier. The principal net upward and downward shifts among the states are 
summarized for each decade in the 1870-1910 period in Table 42. 



Table 42. Principal net shifts in mining labor force among states, by decade, 
1870-1910 



1870-1880 


1880-1890 


1890-1900 


1900-1910 








Net u 


pward shift 








Colo. 


44.40 


Ohio 


21.94 


Wash. 


15.18 


W. Va. 


31.18 


Pa. 


11.49 


Pa. 


10.51 


W. Va. 


13.81 


Pa. 


19.31 


Ariz. 


7.49 


Mich. 


9.81 


Ala. 


12.12 


Minn. 


8.67 


N. & S. 
















Daks. 


6.16 


Ala. 


7.93 


Okla. 


9.07 


Nev. 


5.75 


Kans. 


4.72 


Mo. 


7.53 


111. 


6.37 


Ky. 


5.59 


111. 


3.97 


W. Va. 


4.74 


Tenn. 


5.57 


Okla. 


4.81 


Ind. 


3.54 


Tenn. 


4.21 


Minn. 


5.33 


N. Y. 


3.65 


Mich. 


3.19 


Mont. 


3.78 


Ariz. 


4.77 


Ariz. 


3.02 


Utah 


3.05 


Wyo. 


3.07 


Ore. 


3.62 


Ind. 


2.98 


Iowa 


2.62 


Minn. 


3.05 


Pa. 


3.00 


Tex. 


2.97 


Share of 10 














leading 
















states in 
















total net 
















upward 
















shift 


90.63 




76.57 




78.84 




87.93 








Net downward shift 








Calif. 


30.47 


Calif. 


41.77 


Ohio 


15.39 


Colo. 


21.84 


Ohio 


25.12 


Colo. 


28.61 


Calif. 


15.34 


Calif. 


13.66 


Nev. 


10.26 


Nev. 


6.49 


N. Y. 


13.45 


Mont. 


10.12 


Mont. 


9.15 


N. J. 


4.43 


Mich. 


10.20 


Wash. 


9.63 


Idaho 


8.96 


Ariz. 


4.36 


Nev. 


9.39 


Mo. 


9.24 


N. Y. 


3.92 


Ore. 


4.01 


Wis. 


5.08 


Idaho 


5.63 


Ore. 


3.91 


N. & S. 
Daks. 


3.78 


Colo. 


4.57 


Ore. 


5.58 


Conn. 


2.33 


Idaho 


2.50 


Md. 


4.05 


Mich. 


3.54 


N. J. 


1.25 


Md. 


1.29 


Conn. 


3.57 


Iowa 


3.52 


Md. 


1.10 


Conn. 


.68 


Mass. 


2.61 


Ohio 


2.60 


Share of 10 














leading 
















states in 
















total net 
















downward 
















shift 


96.47 




97.92 




83.65 




85.36 



The figures in italics are absolute downward shifts. 
Source: Appendix Table A-3. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 147 

Activity in the extraction of basic metal ores, notably iron and copper, was 
most marked in Michigan (especially in 1870-1880) and Illinois (1880-1910). 
But the most spectacular fluctuations in this branch of the mineral economy 
occurred in the Mountain region. On balance, Colorado, Utah, and Wyoming 
ended up the 1870-1910 period with a net upward shift representing some 7% 
of the total; Montana and Idaho, on the other hand, registered a combined net 
downward shift of more than 13%. Certain decennial fluctuations in the Moun- 
tain region were quite remarkable. Thus Colorado experienced 44% of the total 
net upward shift during the 1870's (see Table 42) and almost 29% of the total 
net downward shift in the following decade. Similarly in the Southwest, Arizona 
showed a net upward shift for the seventies of more that 7%, a net downward 
shift of 4% in the eighties, and a further net upward shift of just less than 5% 
in the last decade of the century. 

The appearance of Oklahoma and Texas among states with sizable net upward 
shifts around the turn of the century signalizes the beginnings of the petroleum 
industry (also of some coal mining), but the greatest relative net increase in the 
Southwest before 1910 was in Arizona (almost 4% over the period as a whole), 
mostly in connection with copper mining. 

New England had little to contribute to the nation's mineral economy at any 
time, and even the modest quarrying activity carried on in some parts was a 
wasting asset through most of the period. 

Those who had dug for precious metals in the Mountain and Far West regions 
during the third quarter of the nineteenth century had set in motion a chain of 
events which had led to the opening up of the entire West. But already by 1890 
it was evident that much of the more enduring mineral wealth of the nation was 
still to be extracted from regions east of the Continental Divide. 7 



FOREST ACTIVITIES 

Throughout the whole period 1870-1910 agricultural and mineral products 
together accounted for 90% or more of the gross value of primary resources 
extracted in the United States. Primary forest products (hardwood and soft- 
wood lumber plus pulpwood but excluding wood for fuel) at no time contributed 
more than 7% of the total, but the share fluctuated less than that of either of 
the two principal resource sectors. The value of primary forest output in current 
dollars increased slowly throughout the entire period, and down until 1910 
forest employment also showed considerable increases over each census decade. 

The regional shifts in forest employment in the late nineteenth and early 
twentieth centuries were perhaps even more dramatic than those in mining 
activity. Almost every section of the country was involved at some time or 

7 W. J. Trimble, The Mining Advance into the Inland Empire (Madison: University of 
Wisconsin Press, 1914), Isaac Lippincott, Economic Development of the United States (New 
York: D. Appleton & Co., 1921), pp. 326-347, lists Pennsylvania, West Virginia, Illinois and 
Ohio as much greater producers of mineral wealth than any other states before World War 
I; next in order came California, Michigan, Arizona, Montana, Oklahoma, Missouri, 
Colorado, Minnesota, Alabama, Texas, and New Mexico. 



148 / Regional Economic Development: 1870-1950 



Table 43. Principal net shifts in forest labor force among states, by decade, 
1870-1910 



1870-1880 


1880-1890 


1890-1900 


1900-1910 








Net u 


pivard shift 








Mich. 


58.63 


Wash. 


22.65 


Ark. 


15.03 


Wash. 


37.56 


Colo. 


5.43 


Pa. 


21.29 


Tenn. 


11.48 


W. Va. 


15.44 


Mo. 


5.26 


N. C. 


10.16 


Wash. 


9.07 


Ore. 


10.09 


Tex. 


4.47 


Ore. 


6.78 


Tex. 


8.14 


La. 


9.21 


Ore. 


2.99 


La. 


5.78 


La. 


7.63 


Miss. 


6.44 


S. C. 


2.96 


Ky. 


5.62 


Fla. 


7.55 


Idaho 


5.53 


Wis. 


2.91 


W. Va. 


5.00 


Ga. 


6.33 


N. C. 


3.68 


Calif. 


2.90 


Ark. 


4.22 


Miss. 


6.14 


Okla. 


3.62 


N. Mex. 


2.14 


Va. 


3.59 


W. Va. 


5.44 


Maine 


2.86 


Mont. 


1.83 


Mont. 


3.25 


Minn. 


4.29 


Tex. 


2.17 


Share of 10 














leading 
















states in 
















total net 
















upward 
















shift 


89.52 




88.34 




81.10 




96.60 








Net downward shift 








Maine 


19.25 


Mich. 


39.88 


Mich. 


29.12 


Pa. 


26.79 


Pa. 


19.03 


Calif. 


19.60 


Pa. 


21.05 


Mich. 


12.41 


N. Y. 


13.29 


N. Y. 


6.36 


Wis. 


14.09 


Ga. 


7.10 


Nev. 


7.09 


Tex. 


5.85 


Colo. 


526 


Wis. 


6.12 


Ind. 


6.27 


Minn. 


4.94 


Calif. 


4.97 


Ark. 


5.23 


La. 


4.10 


Wis. 


4.80 


Maine 


4.84 


Ky. 


5.04 


Tenn. 


4.01 


111. 


3.56 


N. Y. 


4.18 


111. 


4.99 


Ohio 


3.35 


N. H. 


2.80 


Ind. 


3.65 


Ohio 


4.21 


Minn. 


3.16 


Nev. 


2.59 


Nev. 


2.24 


Calif. 


3.26 


Ala. 


3.05 


Iowa 


2.03 


Mont. 


2.08 


Mo. 


2.81 


Share of 10 














leading 
















states in 
















total net 
















downward 
















shift 


82.60 




92.41 




91.48 




77.96 



The figures in italics are absolute downward shifts 
Source: Appendix Table A-4. 



another except the Mountain and Plains regions, where for the most part forest 
cover was minimal. The heaviest net upward shifts occurred in the Pacific 
Northwest and the Southeast. Washington and Oregon together accounted for 
more than 35% of the total upward shift (Figure 26), and West Virginia for 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 149 




mk - 2 .3 



K; : ] Absolute upward shift 
■/] Relative downward shift 
I Absolute downward shift 




V^\ 



Figure 26. Net Shift in Forest Labor Force, 1870-1910. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 48.9. Total net shift as % of 1870 forest labor force: 
213.2. Source : Appendix Table C. 



more than 11% (second only to Washington). There were also sizable upward 
shifts in Louisiana, Arkansas, Mississippi, Florida, and Virginia in the South- 
east. Until 1900 logging activity remained bouyant in some of the Great Lakes 
states also, but thereafter went into rapid decline. During the nineties, Wiscon- 
sin, Michigan, and Minnesota (in the Plains) were the country's leading pro- 
ducers, but by 1910 they were net importers of lumber and their cut-over pine- 
lands were already being abandoned. 8 

The heaviest net downward shift in forest employment came, of course, in the 
states which had provided the bulk of the nation's lumber for nearly three 
centuries — the once heavily forested states of the New England and Middle 
Atlantic regions. Pennsylvania, New York, and Maine accounted for more than 
40% of the total net downward shift between 1870 and 1910. There was also a 
marked decline — nearly 16% — in logging activities in California and Nevada, 
which had been the earliest sources of softwood supply in the Far West 
(Table 43). 



8 R. F. Fries, Empire in Pine: The Story of Lumbering in Wisconsin, 1830-1900 (Madison: 
State Historical Society of Wisconsin, 1951). Also, Isaac Lippincott, op. cit., pp. 348-359. 



150 / Regional Economic Development: 1870-1950 

FISHING ACTIVITIES 

Important as fishing may be locally in a number of coastal areas, its contri- 
bution to total resource value has not exceeded 1% since 1900. In several 
regions, however — the New England, Middle Atlantic, Southeast, and Far West 
regions — there is an historic concern with fishing, which has provided a modest 
livelihood for tens of thousands of people. During the present century, more- 
over, it has supplied a major part of the material on which the growth of the 
important canning and frozen food industries has been based. 

The value of the catch (excluding the whale fishery) rose from around 
$11,000,000 in 1870 to more than $54,000,000 in 1910. 9 The principal inter- 
regional changes in fishing employment over these years (summarized in Figure 
27) involved the growth of the Gulf and Far Western fisheries and the relative 




,olute upword shifl 
I ~£] Relotive downword shift 
i I Absolute downword shift 



Figure 27. Net Shift in Fishing Labor Force, 1870-1910. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 55.6. Total net shift as % of 1870 fishing labor force: 
80.6. Source: Appendix Table C. 



decline of those in New England and, to a lesser extent, the Middle Atlantic 
region. During the seventies, the largest net increases occurred in Oregon, 
California, and Maryland. But in the ensuing decade the Far Western states 
declined relatively and the Chesapeake states of Virginia and Maryland scored 
the greatest relative gains; smaller net upward shifts also occurred in North 
Carolina, Florida, and Louisiana during this decade. By the turn of the century, 

9 This figure is based on data provided in the detailed federal census of fisheries made 
in 1908. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 151 

however, the growth of the California and Gulf fisheries and of some of the 
inland fisheries (notably in Illinois and Arkansas) had produced relative net 
downward shifts in the Chesapeake area. 

Summary 

The occupation of the greater West after 1870 greatly enlarged the area of the 
nation's resource economy. Between 1870 and 1910 the proportion of the pri- 
mary labor force west of the Mississippi increased by more than 84%, and con- 
stituted almost a third of all resource employment in the latter year. By far the 
largest part of this increase was contributed by agriculture. The essential role 
of the West in these years was to augment the flow of raw materials into the 
industrializing Northeast and, in the case of agriculture, to supply international 
markets as well. Apart from farm produce, the West's most important output 
was mineral ores, notably copper, lead, zinc, and, later, iron. Around the turn 
of the century the development of oil resources in California (after 1895), in 
Texas (after 1898), and in Oklahoma (after 1904) represented a major addition 
to the nation's fuel economy. Following the rapid destruction of the Great Lakes 
forests in the late nineteenth century, the Pacific Northwest became an important 
source for the nation's lumber supply. The growth of coal mining in three of 
the four Eastern regions and the spectacular rise of lumbering in the Southeast 
were perhaps the only new developments east of the Mississippi to compare with 
the contributions of the West. 



Manufacturing Activities, 1870-1910 

The most striking feature in the history of American manufacturing is the 
enduring strength of the Northeast. The industrial pre-eminence of the southern 
New England and Middle Atlantic states, of course, dates back to the beginning 
of the nineteenth century; and, after 1850, the Northeast's position was rein- 
forced by the spread of manufacturing into the Great Lakes states. By 1870, 
80% of the nation's manufacturing employment was concentrated in these three 
regions. Thereafter, technological change and the growth of the transportation 
system permitted some wider diffusion of manufacture, notably into the upper 
Southeast and eastern Plains states and, more recently, into the Far West. But 
even today the great industrial belt in the Northeast continues to dominate the 
regional structure of American manufactures much as it did at the beginning of 
this century. 

Table 44 shows the interregional redistribution of the manufacturing labor 
force between 1870 and 1910. The shares of the New England and Middle 
Atlantic regions in the nation's manufacturing employment declined over these 
years, and the shares of all other regions increased. The general diffusion, how- 
ever, abated somewhat after 1890. The Mountain region's share in 1910 was 
only slightly greater than in 1890, while that of the Plains region had actually 
declined. The Southwest and Far West maintained approximately the same 
rates of relative growth between 1890 and 1910 as between 1870 and 1890. 



152 / Regional Economic Development: 1870—1950 

Only the Southeast showed a significantly greater rate of relative increase after 
1890 than before. New England's relative decline was as great in the later of 
these two periods as in the earlier, but the decline in the Middle Atlantic region 
relative to other regions was somewhat slower after 1890 than between 1870 
and 1890. 



Table 44. Regional distribution of the manufacturing labor force, 1870, 1890, 
and 1910 



Region 



1870 



1890 



1910 



United States 



New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 



2,643,417 


5,525,692 


10,656,545 


100% 


100% 


100% 


21.65 


17.38 


13.42 


38.35 


35.84 


33.55 


20.36 


21.23 


22.61 


9.94 


10.42 


12.69 


6.52 


8.88 


8.35 


.72 


1.31 


2.74 


.35 


1.44 


1.64 


2.11 


3.50 


5.00 



Source: Appendix Table A-6. 



Table 45, which extends the data for the three great regions to 1950, shows 
that surprisingly little redistribution of the manufacturing labor force has taken 
place since 1890. The Northeast experienced a relative decline in its share of 
about 9%, the Southeast an increase of some 48% (on a low base), and the 
West an increase of about 12%. The further growth of manufactures, especially 
after 1910, has involved no major regional relocation and only the Southeast 
among the great regions has continued, moderately, to show an uninterrupted 
improvement in its relative position. In the West, relative declines in the shares 
of the Plains and Mountain regions have offset the continuing growth of the 
Southwest and Far West. 



Table 45. Distribution of manufacturing labor force, by great region, 
1870-1950 



Region 



1870 



1890 



1910 



1930 



1950 



United States 

Northeast 
Southeast 
West 



100% 



100% 



100% 



100% 



100% 



80.36 


74.45 


69.58 


66.38 


67.64 


9.94 


10.42 


12.69 


14.38 


15.43 


9.70 


15.13 


17.73 


19.24 


16.93 



Source: Appendix Table A-6. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 153 

Another useful measure of manufacturing activity is provided by value added 
by manufacture. 10 Table 46 shows the regional shares of total value added by 
manufacture for the period 1870-1910. In general, each region's share is roughly 
proportionate to its share of all manufacturing employment. Indeed, the parallel 
is much closer than in the case of resources value and employment. 



Table 46. Regional distribution of U. S. value added by manufacture, 1870, 
1890, and 1910 



Region 1870 1890 1910 

United States 



New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 



$1,577,387* 


$3,453,518* 


$8,188,527* 


100% 


100% 


100% 


23.95 


17.48 


14.32 


42.16 


40.05 


36.90 


18.07 


24.35 


25.59 


6.23 


6.83 


10.04 


6.68 


7.00 


6.42 


.31 


.68 


1.49 


.24 


.68 


1.21 


2.36 


2.93 


4.03 



*(000's of current dollars). 
Source: Lee, Miller, Brainerd, and Easterlin (see note for Table 4), Table M-8. 



Table 47. Distribution of manufacturing labor force and value added by 
manufacture, by great region, 1870, 1890, and 1910 



Region 




1870 


1890 




1910 




United States 


L.F. 

100% 


V.A. 

100% 


L.F. 

100% 


V.A. 
100% 


L.F. 

100% 


V.A. 

100% 


Northeast 
Southeast 
West 


80.36 
9.94 
9.70 


84.18 
6.23 
9.59 


74.45 
10.42 
15.13 


81.88 

6.83 

11.29 


69.58 
12.69 

17.73 


76.81 
10.04 
13.15 



Source: Appendix Table A-6; Table 46. 



10 Unlike the figure for value of extracted resources, value added by manufacture is a 
"net" figure, derived by subtracting the cost of raw materials, semi-manufactured parts and 
components, supplies, fuels, purchased electric energy, and contract work from the value of 
shipments of manufacturing establishments. This measure avoids duplication in the value 
of shipments which results from the use of products of some establishments as materials by 
others. As such, it constitutes the best value measure available for comparing the relative 
economic importance of manufacturing among industries or regions. 



154 / Regional Economic Development: 1870—1950 

As Table 47 and Figure 28 show, the Northeast achieved a persistently higher 
proportion of the nation's value added by manufacture than it did of manu- 
facturing employment. This was true of no other great region in the period 
before 1910. Only after 1890 did the Southeast begin to reduce the disparity 
between its shares of employment and value added, while between 1870 and 
1890 the growth of value added in the West did not keep pace with the growth of 
manufacturing labor force. 

An analysis of shifts in manufacturing employment among the individual 
states for 1870-1910 shows, not surprisingly, the more urban-industrial states 
to have sustained the greatest relative losses and many of the more rural-agricul- 
tural states to have achieved the greatest relative gains. As late as 1870 the 
industrial states of the New England and Middle Atlantic regions contained 60% 
of all manufacturing employment; by 1910 their share had been reduced to 
about 47%. But, considering the growth of manufacturing in other parts of the 
country over these years, the astonishing fact is that a state like Pennsylvania, 
for example, did not show a greater net downward shift — about 4% over the 
four decades (see Figure 29). (In absolute terms, of course, manufacturing 
employment in Pennsylvania actually increased from less than 327,000 in 1870 
to more than 1,250,000 in 1910.) New Jersey actually experienced a relative 
net upward shift of more than 4%. The greatest relative losses were in New 



■ ■ ■■ „ 



1370 



Manufacturing labor Force 

1890 



::;;: ;• 4- ■ J. . . r .' 4.. .44 :.. ; 4 4 4 



; . .;. .■ ; ■. 



5.W. «, MM .0% 

PI. 6.5% 






2,643,417 
'Employ 6 



5,525,692 
Employed j 



10,656,545 
Employed 



• 




Value Added by Manufacture 

ill i ■■ 



S.w.*Wr.t.4% 

PI. 7.0 % 
S.8. 6.8 




.W. 2.9% 

NJE.t7.59b 



$3,453,513,000 



,, 



• 'i Hi 




Figure 28. Regional Distribution of U. S. Manufacturing Labor Force and 
Value Added by Manufacture, 1870, 1890, and 1910. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 155 








Absolute upward shift 
| Relative downward shift 



Figure 29. Net Shift in Manufacturing Labor Force, 1870-1910. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 20.9. Total net shift as % of 1370 manufacturing labor 
force: 63.3. Source: Appendix Table C. 



York and Massachusetts, where the shifts were 24% and 27%, respectively, 
of the total net downward shift. 

Only five states had net increases amounting to more than 5% of the total net 
upward shift. These were: Illinois (13.2%), Washington (8.5%), Minnesota 
(7.1%), Texas (7.1%), and California (6.9%). The principal net upward and 
downward shifts are shown by decades in Table 48. 

To give greater definition to these changes in manufacturing employment, the 
regional shares of employment in ten selected branches of manufactures are 
summarized in Table 49. 11 The shift analysis has shown, for example that 
there was a net upward shift of secondary employment in every Southwestern 
and Mountain state between 1870 and 1910. These same two regions, 
moreover, show growth in more of the ten selected branches of manufacture 
than any of the other regions — in eight out of the ten. Nevertheless, the actual 
numbers employed by 1910 in any of these industries in the Mountain states 
were still only a minute proportion of national employment in each case. In- 
creases were mostly confined to Colorado, and petroleum refining and lumber 



"The basis of selection for these ten manufacturing activities is simply that it has been 
possible to devise fairly consistent and comparable figures on them for the entire period 
since 1870. They are not necessarily the most important industries either in terms of employ- 
ment or value. A comprehensive treatment of recent changes in manufactures is given in 
Part IV of this study. 



156 / Regional Economic Development: 1870—1950 

products were the only industries in which there was any discernible contri- 
bution beyond the possible needs of local population. These same two processing 
industries loomed large in the Southwest at the turn of the century; and in 
Texas and Oklahoma, at least, petroleum refining had risen to almost 10% of 
the total employment in that industry by 1910. Moreover, by reason of its 
relatively large population, Texas also contained most other branches of manu- 



Table 48. Principal net shifts in manufacturing labor force among states, 
1870-1910 



1870-1880 


1880-1890 


1890-1900 


1900-1910 








Net upward shift 








Pa. 


10.24 


111. 


15.37 


111. 


17.37 


Wash. 


13.76 


Calif. 


9.31 


Wash. 


10.93 


N. C. 


10.47 


Calif. 


12.35 


111. 


9.17 


Nebr. 


10.76 


s. c. 


7.99 


Tex. 


10.35 


N. J. 


9.00 


Minn. 


10.63 


Ind. 


7.88 


Okla. 


8.64 


Colo. 


8.88 


Tex. 


5.69 


N. J. 


7.09 


Ore. 


6.43 


Minn. 


8.33 


Colo. 


4.89 


Okla. 


5.91 


La. 


5.64 


Wis. 


6.65 


Wis. 


4.17 


Ala. 


4.96 


Mich. 


5.00 


Nebr. 


6.15 


Ala. 


3.32 


Ohio 


4.68 


Fla. 


4.50 


Kans. 


5.01 


Ga. 


3.26 


Wis. 


3.96 


IU. 


4.40 


Tex. 


4.56 


Ore. 


3.10 


Ga. 


3.82 


Miss. 


2.86 


Share of 10 














leading 
















states in 
















total net 
















upward 
















shift 


77.30 




72.12 




74.13 




73.93 








Net downward shift 








Mass. 


24.90 


N. Y. 


26.41 


Mass. 


20.70 


Mass. 


26.23 


N. Y. 


23.98 


Mass. 


18.65 


N. Y. 


11.16 


N. Y. 


18.22 


Ohio 


13.31 


Pa. 


9.05 


Nebr. 


10.24 


Pa. 


12.76 


Maine 


7.93 


Conn. 


6.51 


Maine 


8.72 


Conn. 


5,27 


N. H. 


4.70 


Maine 


5,73 


N. H. 


7.11 


N. H. 


5.24 


Vt. 


3.33 


Ind. 


4.91 


Md. 


5.22 


Iowa 


5.08 


La. 


3.21 


R. I. 


4.70 


Calif. 


5.06 


R. I. 


4.20 


Md. 


3.01 


N. H. 


4.10 


Conn. 


4.98 


Md. 


4.07 


Va. 


3.00 


Md. 


3.97 


Ky. 


3.65 


Mo. 


3.84 


Tenn. 


2.74 


Ohio 


3.96 


Colo. 


3.50 


Maine 


3.71 


Share of 10 














leading 
















states in 
















total net 
















downward 
















shift 


90.11 




87.99 




80.34 




88.62 



The figures in italics are absolute downward shifts. 
Source: Appendix Table A-6. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 157 

facture, the close association with population being most evident in the case of 
printing and publishing. 

The Far West, also a growth region, showed development in seven of the ten 
manufacturing categories (in this case, nine categories, as the region has virtu- 
ally no cotton textile industry). Most of the general expansion between 1870 
and 1910 was in California where, of course, the population increase was also 
greatest; the modest proportion of manufacturing development before 1910 
suggests that here, as in other regions of the West, industries served local or 
regional markets rather than the nation as a whole. Their growth was the result 
of geographical remoteness from the national centers of manufacturing in the 
Northeast, and they demonstrate the important historical fact that the Far West, 
and California in particular, developed a manufacturing economy stemming in 
no small part from its particular geographic position. 

There is one likely exception to this tendency. The largest single concentra- 
tion of manufacturing employment in the Far West before 1910 was in lumber 
and lumber products; the region had 11.4% of the total national employment in 
the industry. More than half of this regional proportion, moreover, was located 
in the state of Washington, where this industry provided the major manufactur- 
ing export of the region. By 1910 also, more than 23% of the nation's primary 
employment in forestry and logging was located in this region, and since 1880 
Washington had been the principal contributor to the region's forest labor 
force (see Table 43). 

The contribution of the Far West, Southwest, and Mountain regions to the 
growth of the American economy before 1910 appears to have been confined 
almost wholly to the resource sector, notably forests and minerals, together with 
a few departments of agriculture; with the exception of lumber products, the 
local manufacture or processing of these primary materials had so far assumed 
only minor importance in the nation's economy. 

The Great Lakes region, which experienced increases in seven of the ten 
industries, was clearly the most important manufacturing growth area to emerge 
in the period after the Civil War. In 1870, for example, Ohio was already one 
of the leading manufacturing states, and by 1900 it had been joined by Illinois. 
By 1910 the Great Lakes region contributed more than 20% of national employ- 
ment in half of our selected industries. Its furniture and fixtures industries, 
located principally in Michigan and Illinois and based on the region's rich lum- 
ber supply, employed 47% of the nation's total workers in this category by 
1910. Iron and steel production had also become a major activity, especially 
in Ohio and Illinois, and the share of the five Great Lakes states was almost 
one-third of national employment in this basic industry. The region's share of 
employment in paper products and in printing and publishing by 1910 was 
about one-fourth of the national total; Ohio, Wisconsin, and Illinois were the 
leading states in these categories. Not surprisingly, petroleum-refining and lum- 
ber products were both smaller proportions of the national total by 1910 in the 
Great Lakes region than in 1870, but even in these two fields actual employment 
had risen markedly. Of least significance at any time in this region were the 
basic textile industries. There was an absolute decline in woolen and worsted 



158 / Regional Economic Development: 1870-1950 



Table 49. Regional distribution of employment in ten selected manufacturing 
industries, 1870, 1910, and 1954 





Woolen & worsted 


goods 


Cotton & cotton 


goods 


Region 


1870 


1910 


1954 


1870 


1910 


1954 


United States 


92,973 


175,176 


93,285 


135,763 


387,771 


495,829 






(per 


cent) 








New England 


55.17 


63.08 


49.82 


70.00 


49.75 


11.53 


Middle Atlantic 


29.90 


29.64 


21.92 


21.45 


10.21 


6.14 


Great Lakes 


8.59 


3.35 


6.11 


0.78 


0.98 


0.62 


Southeast 


3.00 


2.99 


19.59 


7.28 


38.49 


79.77 


Plains 


2.20 


0.38 


0.55 


0.27 


0.15 


0.21 


Southwest 


0.13 


0.03 


0.12 


0.21 


0.42 


1.65 


Mountain 


0.11 


0.06 


0.09 


0.01 


— 


0.02 


Far West 


0.90 


0.47 


1.80 


— 


— 


0.06 




Iron & steel products 


Petroleum refining 


Region 


1870 


1910 


1954 


1870 


1910 


1954 


United States 


141,003 


952,527 


987,419 


1,747 


16,640 


153,072 






(per 


cent) 








New England 


10.99 


13.11 


3.79 


8.01 


— 


0.71 


Middle Atlantic 


57.60 


42.88 


35.41 


53.18 


56.65 


20.22 


Great Lakes 


18.92 


32.72 


42.56 


31.08 


17.65 


18.18 


Southeast 


8.68 


5.40 


7.85 


6.36 


2.69 


10.91 


Plains 


3.05 


3.18 


2.71 


0.80 


3.46 


4.18 


Southwest 


0.02 


0.50 


1.77 


0.17 


9.97 


30.66 


Mountain 


0.02 


0.45 


1.40 


— 


2.69 


3.34 


Far West 


0.72 


1.76 


4.51 


0.40 


6.89 


11.80 




Leather 


& leather 


goods 


Lumber & lumber 


products 


Region 


1870 


1910 


1954 


1870 


1910 


1954 


United States 


177,963 


295,973 


356,578 


163,637 


784,989 


645,936 






(per 


cent) 








New England 


42.06 


42.17 


30.87 


12.75 


6.10 


5.69 


Middle Atlantic 


34.71 


26.82 


30.27 


24.33 


10.21 


6.75 


Great Lakes 


13.88 


17.90 


16.51 


35.28 


15.85 


10.22 


Southeast 


4.87 


3.58 


7.82 


14.42 


44.77 


38.18 


Plains 


2.97 


8.09 


10.98 


7.66 


5.86 


3.71 


Southwest 


0.16 


0.06 


1.07 


1.12 


4.10 


4.23 


Mountain 


0.16 


0.18 


0.58 


0.73 


1.70 


3.44 


Far West 


1.19 


1.20 


1.90 


3.71 


11.41 


27.78 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 159 



Furniture and fixtures 



Paper and paper products 



Region 


1870 


1910 


1954 


1870 


1910 


1954 


United States 


57,091 


144,140 


340,694 


18,779 


147,808 


530,210 






(per 


cent) 








New England 


24.65 


7.60 


5.86 


38.82 


30.12 


13.27 


Middle Atlantic 


33.89 


26.82 


21.82 


42.33 


37.34 


25.89 


Great Lakes 


29.29 


46.83 


29.98 


14.85 


25.21 


26.05 


Southeast 


4.59 


10.31 


26.40 


2.79 


2.88 


18.91 


Plains 


6.44 


5.36 


4.06 


0.96 


2.74 


5.84 


Southwest 


0.25 


0.60 


3.49 


— 


0.21 


1.65 


Mountain 


0.18 


0.20 


0.39 


0.04 


0.07 


0.36 


Far West 


0.71 


2.28 


8.00 


0.21 


1.43 


8.03 




Printing & publishing 


Tobacco products 


Region 


1870 


1910 


1954 


1870 


1910 


1954 


United States 


30,924 


388,466 


804,386 


47,848 


197,637 


96,240 






(per 


cent) 








New England 


15.38 


8.97 


7.46 


4.49 


3.82 


1.32 


Middle Atlantic 


44.04 


37.79 


34.48 


39.71 


44.96 


22.95 


Great Lakes 


19.26 


24.29 


27.11 


19.30 


20.96 


3.85 


Southeast 


7.72 


7.78 


8.76 


24.03 


21.70 


69.75 


Plains 


9.81 


11.43 


8.73 


8.45 


6.24 


1.24 


Southwest 


0.59 


2.82 


3.92 


0.12 


0.28 


0.52 


Mountain 


0.22 


1.88 


1.35 


— 


0.56 


— 


Far West 


2.98 


5.04 


8.19 


3.90 


1.48 


0.37 



Source: 9th U.S. Census, 1870, Vol. Ill, Statistics of Wealth and Industry, Table VIII (c) 
13th U.S. Census, 1910, Vol. X, Manufactures, "Reports for Principal Industries, 
passim; U.S. Census of Manufactures, 1954, Vol. II, Pts. 1 and 2. 



goods manufacture after 1870, while the region's share of employment in cotton 
goods rose only from about 0.7% to 0.9% of the national total between 1870 
and 1910. Thus throughout this period the Great Lakes states were not only 
processing their indigenous supplies of natural resources but, in iron ore and 
wood pulp, were already drawing upon the materials of other regions, such as 
the Southeast and the Plains. Their proportions of national employment, more- 
over, in certain crucial branches of manufacture indicate that they were already 
supplying a much larger share of the national market than that contained within 
their own borders. 

Two regions, the Plains and the Southeast, experienced moderate manufac- 
turing growth during the period 1870-1910. They more or less broke even in 
their shares of the ten selected industries; both show relative increases in five 
categories and relative decreases in the other five. Only two of the expanding 



160 / Regional Economic Development: 1870-1950 

manufactures in the Plains, however, amounted to any considerable proportion 
of national employment by 1910. Leather and leather products, centered chiefly 
in Missouri and to a lesser extent in Minnesota, accounted for more than 8% 
and printing and publishing for 11.4% of total employment in these categories. 
Much smaller shares were registered by iron and steel products (again chiefly in 
Missouri and Minnesota) and petroleum refining (mostly in Kansas). Our ten 
industries, however, do not include one of the major manufacturing develop- 
ments of the Plains during this period, namely flour and grist mill products. 
Here the eastern Plains states achieved a remarkable expansion as the major 
wheat belts of the country shifted to the westward out of the Great Lakes states. 
By 1910 the Plains region employed more than 27% of all workers in the 
milling industry and was rapidly becoming the major source of the nation's 
supply of flour and grist mill products. 

The leading edge of growth in the Southeast, even before the end of Recon- 
struction, was provided by lumber products and, to a lesser extent, by cotton 
textiles. By 1910 this region accounted for nearly 45% of total employment in 
lumber products and almost 39% of all cotton textiles employment. All South- 
eastern states seem to have participated in the development of lumber products, 
but cotton textile growth was centered almost exclusively in the Carolinas, 
Georgia, and Alabama. The furniture and fixtures industry, mostly in North 
Carolina and Tennessee, also contributed to the growth of Southern manufac- 
tures; the region had more than 10% of national employment in this industry 
by 1910. Minor advances in petroleum-refining had been made in Louisiana by 
1910 also and in paper products in Virginia and West Virginia. In spite of a 
sizable development in iron and steel products in Alabama, Tennessee, and 
Kentucky during this period, however, the Southeast contributed a somewhat 
smaller share of all employment in this crucial sector by 1910 than it had during 
the height of the Reconstruction period. In other words, actual development of 
the iron and steel industry in the Southeast had not kept pace with developments 
elsewhere in the nation. 12 More suggestive, if not more significant, perhaps, is 
the fact that printing and publishing remained a relatively static industry in one 
of the most thickly populated regions of the country. 

Turning, finally, to what in the late nineteenth century was still the heartland 
of American manufactures — the New England and Middle Atlantic regions — we 
find more of the ten selected industries in relative decline than in relative 
growth. New England registered declines in seven and growth in three of these 
industries between 1870 and 1910, and the Middle Atlantic region declined in 
eight and grew in only two. Though this conforms to the analysis of shifts in 
all manufacturing employment shown in Figure 29, there were actually only two 
cases of absolute decline among the ten manufactures and both occurred in New 
England: petroleum-refining had disappeared, or at least did not warrant inclu- 
sion in the Census of Manufactures, and furniture and fixtures production had 

12 C. Vann Woodward, Origins of the New South, 1877-1913, Vol. 9. of A History of the 
South (Baton Rouge: Louisiana State University Press, 1951), pp. 291-320; H. L. Herring, 
Southern Industry and Regional Development (Chapel Hill: University of North Carolina 
Press, 1940), pp. 31-5. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 161 

declined absolutely in every New England state except Vermont. New England's 
three growth industries during this period were woolen and worsted goods in 
Massachusetts and Rhode Island, iron and steel products in Massachusetts and 
Connecticut, and leather products, more especially the boot and shoe manu- 
factures of Massachusetts. In all three of these categories, however, the region 
was not merely processing indigenous materials but by 1890 was already 
importing much of the raw product from other parts of the country. 

In the Middle Atlantic region there were, of course, absolute increases in 
numbers employed in each of the ten selected branches of manufactures; but 
growth, i.e., relative increase in the region's share of national employment, was 
confined to petroleum-refining (chiefly in New Jersey and Pennsylvania) and 
tobacco products (chiefly in Pennsylvania and New York). Pennsylvania, of 
course, had been the birthplace of the great petroleum industry, while New 
Jersey had acquired its later prominence owing to its seaboard location and its 
proximity to the most densely populated urban centers in the country. The 
tobacco industry, however, is a somewhat special case. Before the widespread 
adoption of cigarette smoking and the consequent mechanization of tobacco 
manufactures, employment in this industry was closely associated with popula- 
tion, not unlike printing and publishing or the major food industries. By 1910 
the Middle Atlantic region was by far the largest employer, accounting for 
nearly 45% of all tobacco employment, while the Southeast's share had fallen 
from 24% to 21% since 1870. Within the Southeast, Florida employed a con- 
siderably larger number of tobacco workers than either North Carolina or 
Virginia, owing to the concentration of cigar-making in the Tampa area. The 
growth of cigarette smoking and the mechanization of cigarette manufacture was 
by 1950 to restore all but 30% of tobacco employment to the Southeast and 
largely concentrate it in North Carolina, Virginia, and Kentucky. 

Service Activities in the Regions, 1870-1910 

The tertiary or service sector of the economy has often been regarded as itself 
a key measure of economic development and material progress. As technical 
and organizational advances in the primary (resource) and secondary (manu- 
facturing) sectors have augmented labor productivity, a growing proportion of 
the labor force has been freed from resource extraction and manufactures to 
engage in so-called "services" to business and consumers, e.g., in transportation 
and communications, wholesale and retail trade, finance, recreation, amuse- 
ments, the professions, and other services to individuals. There is, moreover, 
some evidence that when personal incomes rise, persons tend to purchase larger 
proportions of service items than when their personal incomes are lower; that is, 
most tertiary items have higher income elasticity of demand than most primary 
or secondary items. There is, likewise, some correlation between the growth of 
service activities and the degree of urbanization of a population. 13 

13 E. E. Lampard, "History of Cities in the Economically-Advanced Areas," Economic 
Development and Cultural Change, Vol. 3, (January 1955), pp. 99-102; Colin Clark "The 
Economic Functions of a City in Relation to its Size,"/ Econometrica, Vol. 13, (April 1945), 
pp. 97-8. 



162 / Regional Economic Development: 1870-1950 

When the service sector of the economy is defined as "all other activities" — 
i.e., as the total labor force less the workers engaged in primary and secondary 
activities — so many different types of activities are involved that it does not 
seem wise to generalize too freely about their character and significance. 14 But 
with this reservation in mind, let us examine the distribution of tertiary employ- 
ment much as we have previously examined the other two labor force groupings. 
Table 50 shows the changing proportions of all service activities in the eight 
regions and three greater regions of the United States between 1870 and 1910. 
The principal features of these regional changes are the relative stability of the 



Table 50. Regional distribution of services labor force, 1870, 1890, and 1910 



Region 


1870 


1890 


1910 


United States 


3,178,287 


7,345,373 


13,913,976 




100% 


100% 


100% 


New England 


11.45 


9.34 


8.43 


Middle Atlantic 


36.14 


31.69 


29.48 


Great Lakes 


19.95 


20.77 


20.15 


Southeast 


18.55 


16.13 


16.13 


Plains 


8.25 


12.34 


11.84 


Southwest 


1.61 


2.76 


4.72 


Mountain 


.58 


2.14 


2.48 


Far West 


3.47 


4.83 


6.77 


Northeast 


67.54 


61.80 


58.06 


Southeast 


18.55 


16.13 


16.13 


West 


13.91 


22.07 


25.81 



Source: Appendix Table A-7. 



Southeast and some partial reversal in the relative positions of the Northeast and 
the West. (These tendencies, as we shall see later, continue to operate in the 
subsequent period, 1910-1950.) At first the relative losses of the two North- 
eastern seaboard regions were recouped by the central portion of the country, 
notably the Great Lakes and Plains regions, but even before 1910 the Plains had 
failed to sustain its relative growth and the gradual westward shift of tertiary 
activities had continued into the Far West and Southwest. 

Table 51 compares percentage changes in regional shares of the nation's total 
tertiary employment with percentage changes in total labor force and total 
population shares between 1870 and 1910. During these years, which saw the 
flood tide of the westward movement of population, the Great Lakes region was 
the only region east of the Mississippi that increased its share of service employ- 
ment. In contrast, all regions west of the Mississippi increased theirs. The 
Plains, Mountain, and Southwest regions actually increased their shares of 

14 See G. J. Stigler, Trends in Employment in the Service Industries (Princeton: Princeton 
University Press, 1956). 



Changing Regional Structure of U. S. Economy,, 1870-1910 (2) / 163 



Table 51. Percentage change in regional shares of population, total labor force, 
and services labor force, 1870 - 1910 









Services 


Region 


Population 


Labor force 


labor force 


New England 


-18.72 


-26.47 


-26.38 


Middle Atlantic 


-7.04 


-6.62 


- 18.43 


Great Lakes 


- 13.44 


-13.15 


+ 1.00 


Southeast 


-17.85 


-12.59 


-13.05 


Plains 


+ 30.55 


+ 26.19 


+ 43.52 


Southwest 


+ 160.63 


+ 184.40 


+ 193.17 


Mountain 


+ 411.63 


+ 292.98 


+327.59 


Far West 


+ 158.33 


+ 111.84 


+ 95.10 



Source: 17th U. S. Census, 1950, see supra, Table 1, note; Appendix Tables A-l, A-7. 



service employment considerably more than their shares of total labor force. 
The Far West, on the other hand, had a smaller relative increase in service 
employment than in labor force. This disparity is accounted for in part by the 
importance of transport and communications in the three "corridor" regions. 

No two regions in the eastern half of the country show like patterns of change. 
In the Great Lakes region, service employment increased relatively, though 
slightly, while both population and labor force shares declined considerably. In 
the Middle Atlantic region, there was a greater relative decline in the service 
sector than in either labor force or population. In the Southeast, the relative loss 
was greater for population than for labor force or service employment. In New 
England, it was greater for labor force and service employment than for popu- 
lation. 

In view of the gross character of our service sector, an analysis of these shifts 



Table 52. Principal net shifts in services labor force among states, 1870 - 1910 





Net upward shift 




Net downward shift 


Texas 


11.27 


New York 


20.38 


Washington 


10.32 


Pennsylvania 


11.76 


Illinois 


9.86 


Maryland 


8.64 


Minnesota 


9.86 


Virginia 


7.35 


Oklahoma 


7.51 


Kentucky 


6.51 


California 


7.17 


Massachusetts 


6.47 


Colorado 


5.84 


Maine 


5.89 


Dakotas 


5.42 


Ohio 


5.00 


Nebraska 


5.20 


Louisiana 


4.59 


Kansas 


4.81 


Michigan 


4.14 



Source: Appendix Table A-7. 



164 / Regional Economic Development: 1870-1950 

in the detail of particular decades would have little or no significance. Accord- 
ingly, we shall show only the principal net upward and downward shifts in 
service employment over the 1870-1910 period as a whole (Table 52). The 
figures for the states bear a close resemblance to the broad regional changes 
outlined above. Net downward shifts in services are registered chiefly in the 
eastern half of the country and relative net upward shifts in all of the Western 
states except Missouri and Nevada. The westward shifts in services, as in popu- 
lation, were generally greater during decades of prosperity than of depression, 
and reached their greatest volume after 1900. As early as 1880, however, there 
was a relative net upward shift into parts of the Southeast, principally into 
Florida, which, except for West Virginia, was the only Southern state that had 
a net upward shift in service employment over every decade of the 1870-1910 
period. Considering the high proportion of all service employment made up at 
this time of transportation, trading, and domestic services (our service sector 
also includes construction workers), it is clear that the opening up of the West 
was the dominant factor influencing the redistribution of tertiary employment. 
In this respect, the tertiary sector is not very different from either the resource 
or the manufacturing sectors; but in the degree of redistribution, it is affected 
much more profoundly than manufactures. There was less diffusion of manufac- 
tures than of services; hence there was less disparity among the regions in 
services by 1910. 

It would seem that in a growing economy like that of the United States, 
regions do not have to develop any particular stage or level of manufactures in 
order to obtain a fairly wide range of service employments. Services are not 
necessarily dependent on an economic base in manufactures and may indeed 
contribute a large part of a locality's base on their own account — e.g., wholesale 
trade in the Far West or transportation in the Mountain and Plains regions. 
Nevertheless, the fact that Texas, Washington, and California were the leading 
growth states in services west of the Mississippi before 1910 suggests that the 
development of some services is probably contingent on the growth of some 



Table 53. Percentage growth in total labor force, services labor force, and labor 
force in four selected service industries, 1870-1910 



1870-1910 



All labor force +205.19 

All service labor force +337.78 

Transport and communications +506.84 

Trade and finance +377.73 

Professional services +385.93 

Personal and domestic service +67.73 

Source: 9th U.S. Census, 1870, Vol. I; 13th U.S. Census, 1910, Vol. IV, Occupational 

Statistics. Appendix Tables A-l, A-7. 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 165 

types of manufactures. The fact that the principal net downward shifts in serv- 
ice employment occurred in states with already high levels of manufacturing and 
other business activity does not seem to affect the point one way or the other. 
The Middle Atlantic states alone contributed more than 40% of the total net 
downward shift in services in this period and only 34% of the total net down- 
ward shift in manufacturing employment. 

Up to this point we have treated changes in the tertiary sector as a whole 
without attempting any close analysis of their significance because of the hetero- 
geneous nature of the sector. It is doubtful that anything certain can be said 
about so diverse an occupational grouping as a whole. 15 But some definition can 
be given to the changes in this sector by examining certain of the major service 
activities separately. We have selected the four following service activities for 
such an examination for the 1870-1910 period: transportation and communica- 
tions, trade and finance, professional services, and personal and domestic serv- 
ices. As Table 53 shows, the first three of these follow the general tendency of 
service employment to rise faster than the labor force as a whole. The fourth — 
personal and domestic services — shows a different pattern. Indeed, after 1910, 
the numbers in this category go into absolute decline. 16 . 

Transportation and communications showed the fastest growth between 1870 
and 1910. The rates of growth for trade and finance and professional services 
closely paralleled each other during this period. There was marked regional 
differentiation, however, in both the direction and extent of these changes (see 
Table 54). Broadly speaking, the growth in the West was at the expense of the 



Table 54. Regional distribution of labor force in four selected service indus- 
tries, 1870 and 1910 





Transport and 


Trade 


and 


Professional 


Personal 


and 




communications 


finance 


services 


domestic i 


service 


Region 


1870 


1910 


1870 


1910 


1870 


1910 


1870 


1910 


United States 


100% 


100% 


100% 


100% 


100% 


100% 


100% 


100% 


Northeast 


15.26 


7.17 


12.60 


8.58 


12.12 


8.48 


9.58 


8.28 


Middle Atlantic 


35.68 


27.07 


38.60 


29.66 


28.89 


26.44 


33.95 


29.03 


Great Lakes 


20.98 


20.92 


20.54 


21.27 


24.59 


21.10 


20.64 


17.61 


Southeast 


12.71 


15.33 


14.09 


13.78 


18.31 


14.67 


22.11 


22.20 


Plains 


9.99 


13.19 


8.84 


12.75 


10.92 


13.98 


8.08 


10.08 


Southwest 


1.14 


5.10 


1.34 


4.94 


2.01 


5.41 


1.68 


4.75 


Mountain 


.82 


3.45 


.60 


2.21 


.36 


2.79 


.50 


2.09 


Far West 


3.42 


7.77 


3.39 


6.81 


2.80 


7.13 


3.46 


5.96 



Source: See Table 53. 



15 To give only a few examples, it includes transportation workers, dentists, army officers, 
civil servants, housemaids, shopkeepers, real estate brokers, traveling salesmen, many types 
of construction workers, fortune-tellers, Indian guides, and scavengers. 

16 It is personal services inside rather than outside the home that contribute to the decline. 



166 / Regional Economic Development: 1870-1950 

relative position of the Northeast and, to a lesser extent, of the Southeast. The 
Southeast actually increased its relative strength in transport and communica- 
tions and retained an unchanging share of personal and domestic services, but 
it lost ground relatively in trade and finance and in professional services. In the 
Northeast, New England suffered the greatest relative declines except in personal 
and domestic services. Its share of transport and communications was reduced 
by one-half, and its share of both trade and professional services by about one- 
third. Considering the relative size of the Middle Atlantic region in 1870 (it had 
more than a third of all employment in each category except professional serv- 
ices), its losses were fairly moderate. The Great Lakes region actually held its 
own in the transport and trade categories, although it declined relatively in the 
other two. In the period before 1910 the West advanced substantially in all 
categories; in its least buoyant category, personal and domestic services, it 
gained more than 67%. The Southwest and Mountain regions, the least devel- 
oped regions in 1870, made the greatest relative gains; but in absolute terms, of 
course, the development of the Plains and the Far West was more remarkable. 

Table 55 shows the percentage of the labor force that each of these service 
categories comprised in the eight regions and in the nation as a whole in 1870, 
1910, and 1950. All of them except personal and domestic services formed a 
larger proportion of each region's labor force in 1950 than in 1870. The same, 
of course, was true for the nation as a whole; economic growth from 1870 to 
1910 was accompanied by a significant rise in certain types of services, more 
especially services to commerce and industry. That the growth of such services 
was geographically uneven is attested by the fact that two regions, the Southeast 
and Southwest, were below the national average in all four categories both at 
the beginning and at the end of the period. In this sense, these two regions may 
be said to have been "underserviced" by the national standards of the day. Two 
other regions, the Middle Atlantic and the Far West, were above the national 
average in the four selected categories in 1910, as earlier in 1870. 

Table 56 summarizes the percentage growth or decline of each of the four 
service categories as proportions of their total regional labor forces and com- 
pares regional changes in this regard with those for the nation as a whole. As 
already noted, three of the categories were generally growing during this period 
and only one was generally in decline. It is significant that transportation and 
communications services were the most buoyant. Even in the two Northeastern 
seaboard regions, where transport services failed to grow at as fast a rate as in 
the nation as a whole, these services formed a larger part of the regional labor 
forces in 1910 than in 1870. Trade and finance likewise increased their shares 
of the labor forces in every region during this period, although in four of the 
regions the increase was less than that for the nation. All regions also increased 
the professional services contingent in their labor forces; three regions grew 
faster than the nation but of these only the Mountain and Far West regions 
exceeded the national rate of growth in any notable degree. Finally in the one 
declining category, personal and domestic services, four regions declined at a 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 167 



o 
io 

On 
r— I 

m 

c 

cd 



a 

o 

S-H 

J3 



O 
03 



cd 

■4-1 

5 






o 
cd 






O .5 

Q 



a & 



3 «3 

s-l 

H °3 



H °3 



£ o o 



03 °0 



o .a 



H =3 



* tt S 



N Q\ i — i Q\ 0\ 0\ CO ^f 
Oi ^O C ^ lO M_ O N 
tJ LO rjl ts ^ N lO lO 



CoKJNO^t^COCO 



COCMfOCN'-OCNCMCNl 



iflCONifiNNO\N 



CiCNOCOCOr-O-s 



rji^^N^cO^^O 



Ci —■ ! ?i ui ci r^ On IN 



CO'-JCiNiNOKVO 

■^ONvqcooovq^oco 

\ONNrJNOCici 



N N f Ci N N lO N iv. 


cJ 


O 
> 


ft^ «5 ft^ h crj eg m co 


bfi 

CCi 

CD 


© 

LO 
ON 



H N CO 

I> co CO 

io ro io co *6 cd 



CO 

03 


"2 s 


CO 






CO 


C3 r£ 


<v 






CO 


Tjjxj 




"S 




-cs 


w ^ 


J 


CO 

03 


CO 


03 
"S 




cd 

o 


3 
C 

CO 





^ a 

CO S 



« H 



•- 1 co 



168 / Regional Economic Development: 1870—1950 



Table 56. Percentage change in regional distribution of labor force in four 
selected service industries, 1870 - 1910 





Transp. and 


Trade and 




Personal and 


Region 


communication 


finance 


Professional 


domestic 




% 


% 


% 


% 


United States 


+ 98.56 


+ 56.53 


+ 59.12 


- 45.05 


New England 


+ 27.06 


+ 44.96 


+ 51.72 


- 35.44 


Middle Atlantic 


+ 61.38 


+ 28.72 


+ 56.23 


- 49.71 


Great Lakes 


+ 128.23 


+ 86.77 


+ 57.65 


- 45.98 


Southeast 


+ 173.42 


+ 75.33 


+ 45.81 


- 36.88 


Plains 


+ 108.53 


+ 79.24 


+ 61.42 


- 45.58 


Southwest 


+ 214.36 


+ 103.77 


+ 50.59 


- 45.23 


Mountain 


+ 110.87 


+ 46.94 


+ 212.64 


- 41.72 


Far West 


+ 113.17 


+ 48.33 


+ 91.69 


- 55.40 



Figures in italics are above the national average. 
Source: See Table 55. 



rate faster than the national rate, but of these only the Far West greatly exceeded 
the national rate of decline. 

These trends in the growth and redistribution of service activity during the 
late nineteenth century were an important feature of the nation's economic 
growth. They underline the fact that, in order to develop the full resource 
potential of the greater West and integrate Western materials into the nation's 
resource economy, it was necessary to organize a nation-wide system of trans- 
portation and communications together with a whole edifice of related services 
to trade and finance. These services, in turn, help account for the surprisingly 
high levels of urbanization obtaining in some of the Western states before 1910 
without a corresponding growth in local manufactures. Specialized organiza- 
tional functions, therefore, are not merely a way of absorbing excess manpower 
released from more "fundamental" primary and secondary production through 
the introduction of labor-saving techniques; they are an essential concomitant 
of efforts to control and utilize raw materials over wider expanses of conti- 
nental territory. Without these services and facilities, which provided a major 
outlet for new investment at the time, the natural resources of the West could 
not have been made available to the industrializing areas at home and abroad. 
They are evidence of structural changes in the labor force which accompanied 
the "nationalizing" of the American economy. 

Industrialization of the labor force after 1870 did not mean that a majority 
of the gainfully employed became manufacturing wage-earners. Of the total 
labor force of some 38,167,000 in 1910, all categories of skilled, semi-skilled, 
and unskilled workers outside agriculture comprised little more than 39% of 
the total. Farm proprietors (6,000,000), dealers in wholesale and retail trade 
(2,600,000), and other owners, managers, or officials of business enterprises 



Changing Regional Structure of U. S. Economy, 1870-1910 (2) / 169 

comprised almost 23% of the labor force. There were at least 5,500,000 farm 
laborers and an additional 8,000,000 workers in clerical jobs and in services of 
a personal or professional nature. 17 And as the twentieth century progressed, 
manufacturing workers formed an even smaller share of the total labor force. 



"In 1910 nearly 2 million children 15 years old or under held jobs, compared to only 
750,000 in 1870. The number of employed women more than doubled between 1890 and 
1910, exceeding 8 millions (21.2% of the labor force in 1910: about 30% were in domestic 
service) . 



J$ / Changing regional structure of the 
U. S. economy, 1870-1910: 

(3) REGIONAL INDUSTRIAL 
STRUCTURES AND INCOME 



The great advances in industrialization that occurred in the closing decades of 
the nineteenth century profoundly affected economic developments in the various 
parts of the country. Between 1890 and 1910 the economic boundaries came to 
coincide more exactly with the political boundaries of the country than at any 
time since 1803. The completion of the railroad network, the construction of 
oil pipelines, the development of "instantaneous communication" by telegraph 
and telephone, and the beginnings of automobile transportation now linked all 
outlying regions with the historical growth centers of the economy. Henceforth, 
major economic decisions must be taken in the light of truly nation-wide oppor- 
tunities. The full resource potential of every region was now at the disposal of 
a national system of production and could be utilized to meet the growing 
demand of a nation-wide system of distribution and markets. 1 

The consequences of the nation's economic growth for regional development 
can best be understood by comparing regional economic structures with the 
economic structure of the nation as a whole. The changing structure of pro- 
ductive activities reflected in the national labor force between 1870 and 1910 
is shown in Table 57. The proportions of various activities in the national labor 
force serve as a yardstick against which to measure particular developments in 
the regions. For example, in Figure 30, the proportion which the agricultural 
component comprises of each region's labor force is compared with that of the 
national labor force for the 1870-1910 period. The Southeast, Southwest, and 
Plains regions have greater concentrations of agricultural labor than the nation 

1 In 1914, before the outbreak of the World War, G. B. Hotchkiss points out, "the whole 
conception of the domestic market had changed. It was no longer a definite place, but 
people. The manufacturer was able to consider as his market all potential customers, 
wherever located, since he had a means of reaching them with his products and with 
information about his products." See Milestones of Marketing (New York: The Macmillan 
Co., 1938), pp. 220-21. The growth of nation-wide advertising dates from about 1890. 

170 / 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 171 

as a whole throughout this period and may be identified as agricultural regions. 
The Great Lakes region begins the period as a predominantly agricultural 
region, but by 1390 other activities have assumed greater importance and the 
region rapidly becomes less agricultural than the nation as a whole. In the New 
England, Middle Atlantic, Mountain, and Far West regions the concentrations 
of agriculture are persistently below the national average over these years. 



Ut Cent 




Figure 30. Change in the Regional and National Agricultural Labor Force as 
Percentage of Total Labor Force, 1870-1910. 
Source: Appendix Tables A-l and A-2. 



172 / Regional Economic Development: 1870-1950 



Table 57. U. S. labor force, distribution by industry, 1870 - 1910 













(per 


cent of total) 




Agri- 








Manu- 




Year 


culture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


51.48 


1.49 


0.26 


0.22 


21.14 


25.41 


1880 


49.39 


1.71 


0.32 


0.24 


22.09 


26.25 


1890 


40.62 


1.97 


0.54 


0.26 


24.30 


32.31 


1900 


38.83 


2.39 


0.48 


0.24 


24.76 


33.30 


1910 


32.46 


2.53 


0.45 


0.18 


27.92 


36.46 



Source: Appendix Tables A1-A7. 



Industrial Structure of the New England Labor Force, 1870-1910 

The industrial structure of the New England labor force in the 1870-1910 
period, summarized in Table 58, shows the region's economy to be based on 
manufacturing and services rather than indigenous natural resources. Over these 
four decades the agricultural component declined by almost two-thirds, and the 
mining component was negligible. The upsurge in the forest component during 
the first decade of this century was related to the adoption of new woodpulping 
processes which permitted use of some of the region's remaining stands of 
timber in outlying places (notably in Maine). The fisheries, though more con- 
centrated here than in the nation at large, were a declining share of the total 
regional labor force. New England was the most heavily concentrated manu- 
facturing region in the American economy during this period. Unlike the other 
manufacturing regions, its labor force in services was a smaller share of the total 



Table 58. Distribution by industry of New England region labor force, 
1870-1910 













(per 


cent of total) 




Agri- 








Manu- 




Year 


culture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


26.38 


0.41 


0.27 


0.87 


44.05 


28.02 


1880 


21.46 


0.46 


0.23 


0.81 


47.98 


29.06 


1890 


16.37 


0.47 


0.34 


0.76 


47.87 


34.19 


1900 


13.18 


0.34 


0.29 


0.53 


48.16 


37.50 


1910 


9.65 


0.31 


0.31 


0.43 


49.05 


40.25 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 173 

than the labor force in manufactures. The growth of services after 1880, how- 
ever, proceeded at a faster rate than the growth of manufactures, and the earlier 
disparity between the two sectors was much reduced by 1900. 

The differentiation among the states that compose a region are sometimes no 
less marked than the regional differentiation of production activities within the 
national economy. The northern New England states — Vermont, New Hamp- 
shire, and Maine — differ sharply from the southern New England states — Mas- 
sachusetts, Rhode Island, and Connecticut — in their economic structure. The 
extreme case here is Vermont which, over the 1870-1910 period, was always 
more agricultural than the nation as a whole. During the eighties, moreover, 
quarrying in Vermont brought the mining component of its labor force into 
parity with that of the nation as a whole. New Hampshire had a larger propor- 
tion of manufacturing employment throughout these years than either of its 
northern neighbors; it was also somewhat less agricultural and much less of a 
services state. In southern New England, Rhode Island was (and, in fact, still 
is) the most heavily concentrated manufacturing state in the nation. Connecti- 
cut was more agricultural than either Rhode Island or Massachusetts. After 
1900, however, it developed a heavier concentration of manufactures than 
Massachusetts. But Massachusetts had a larger proportion of its labor force in 
service occupations than either Connecticut or Rhode Island. 



Industrial Structure of the Middle Atlantic Labor Force, 1870-1910 

The industrial structure of the labor force in the Middle Atlantic states, sum- 
marized in Table 59, shows a regional economy based on manufactures and 
services with an increasing segment of the labor force engaged in mining activ- 
ities. The mining activities, however, in spite of their vital contribution to the 
region's industrial fuel supply, are not large enough for the region as a whole to 
be identified with mining. Nevertheless, Pennsylvania is one of the great mining 



Table 59. Distribution by industry of Middle Atlantic region labor force, 
1870-1910 













(per 


cent of total) 




Agri- 








Manu- 




Year 


culture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


29.09 


1.97 


0.21 


0.19 


32.13 


36.41 


1880 


24.02 


2.47 


0.19 


0.26 


34.87 


38.19 


1890 


18.39 


2.91 


0.38 


0.34 


35.84 


42.14 


1900 


14.76 


3.49 


0.25 


0.28 


37.28 


43.94 


1910 


10.39 


3.94 


0.15 


0.16 


39.75 


45.61 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



174 / Regional Economic Development: 1870-1950 

states of the American economy: miners comprised almost 5% of its labor force 
in 1870, 7.27% in 1890, and nearly 10.5% in 1910. Maryland is the only other 
state in this region with any significant amount of mining employment. 

The region's agricultural labor force declined by almost two-thirds between 
1870 and 1910, and forestry accounted for an even smaller proportion of total 
employment than the fisheries, which were important locally in some of the 
seaboard states. The concentration of employment in service activities, in excess 
of manufacturing employment, is due in part to the fact that many financial and 
business services, located in New York, New Jersey, and Pennsylvania, actually 
service wide areas of the economy outside the Middle Atlantic region itself. The 
region also includes the District of Columbia, where the service component is 
very large. Service employment in most of the other Middle Atlantic states also 
was a greater proportion of total employment than in the states in other manu- 
facturing regions (except perhaps Massachusetts in New England and Illinois in 
the Great Lakes regions). Delaware was the most agricultural of the Middle 
Atlantic states throughout the 1870-1910 period, though it failed to reach the 
national proportion in this sector. New Jersey was the least agricultural, and 
Delaware and New York had the least mining. Maryland had heavy local con- 
centrations of fisheries and Pennsylvania the greatest forest activity. New 
Jersey, New York, and Pennsylvania, of course, had the greatest concentrations 
in manufactures. 



Industrial Structure of the Labor Force in the Great Lakes Region, 1870-1910 

The industrial structure of the Great Lakes region in the 1870-1910 period is 
shown in Table 60. During these years this region experienced the full impact 
of the industrial revolution, although Ohio had already developed a major con- 
centration of manufactures before 1870. During the nineties the region also 
began to show its potential strength in service activities; here, Illinois was the 



Table 60. Distribution by industry of Great Lakes region labor force, 
1870-1910 













(per 


cent of total) 




Agri- 








Manu- 




Year 


culture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


55.72 


1.13 


0.28 


0.08 


19.65 


23.14 


1880 


50.75 


1.07 


0.52 


0.07 


22.01 


25.58 


1890 


39.59 


2.04 


0.68 


0.10 


25.04 


32.55 


1900 


34.18 


2.31 


0.43 


0.10 


27.28 


35.70 


1910 


25.16 


2.61 


0.31 


0.10 


33.19 


38.63 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 175 

largest contributor both in relative and absolute terms. Until the nineties, how- 
ever, the Great Lakes region was primarily an agricultural region. It was only 
after 1900 that the manufacturing component actually surpassed the agricultural 
component of the labor force. Indiana and Wisconsin remained the most agri- 
cultural states in this region; as late as 1910, one-third of their workers were in 
agricultural pursuits. 

After 1880 the great mineral wealth of the region (in coal, iron, and copper) 
made a crucial contribution to its development. Mining employment in the 
region as a whole exceeded the national average both in 1890 and in 1910 and fell 
only slightly under it in 1900. Except during the depression in the late seventies, 
Ohio always exceeded the national proportion, as did Michigan and Illinois 
after 1890. No Great Lakes state, however, achieved the level of mining concen- 
tration that developed farther east in Pennsylvania or West Virginia. The 
largest concentration occurred in Michigan, which had 3.28% of its labor force 
engaged in mining around 1910. The region's employment in forestry and 
logging activity surpassed the national level from the mid-seventies until about 
1890, when it reached its peak. At this time Michigan had one of the heaviest 
concentrations of forest labor anywhere in the nation, and Wisconsin also 
ranked high. Finally, though the region's stake in fisheries was a minor one, 
Michigan's lake fishing corresponded closely with the national level of fishing 
activity until 1880. 

Clearly, the Great Lakes region was one of the greatest resource regions in the 
American economy. Its mines, fields, and forests made a key contribution to 
national economic growth in the late nineteenth century. Its natural abundance 
not only provided a firm foundation for its own burgeoning manufactures and 
commerce, but also gave new strength to the continuing thrust of industry 
elsewhere in the greater Northeast. By the end of the century it was already 
drawing in natural resources from adjacent states in the Southeast and Plains 
regions. 



Industrial Structure of the Labor Force in the Southeast, 1870-1910 

The industrial structure of the Southeast's labor force in the 1870-1910 period, 
shown in Table 61, contrasts sharply with that of the Great Lakes region. 
Although the Southeast had begun to develop its abundant natural resources and 
diversify its economic base by 1890, the progress made was inadequate in view 
of the large population in this region. As a result, the Southeast fell behind 
the rest of the country in many important respects. For example, the share of 
agricultural employment in its labor force declined only 22% during these years 
as compared with 37% in the nation as a whole and 55% in the Great Lakes 
region. The character and organization of farming through much of the South- 
east at this time suggests that there was severe under-employment in agriculture. 

Somewhat paradoxically, the share of manufacturing in the labor force 
increased faster in the Southeast than in any other region over these years; the 
growth of industry in the "New South" during the eighties, in fact, had become 



176 / Regional Economic Development: 1870-1950 

Table 61. Distribution by industry of Southeast region labor force, 1870-1910 

(per cent of total) 





Agri- 








Manu- 




Year 


culture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


75.13 


0.22 


0.16 


0.19 


7.49 


16.81 


1880 


75.35 


0.31 


0.18 


0.16 


7.81 


16.19 


1890 


66.28 


0.82 


0.48 


0.28 


10.51 


21.63 


1900 


64.95 


1.27 


0.64 


0.29 


11.34 


21.51 


1910 


58.93 


1.73 


0.65 


0.23 


14.46 


24.00 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



a matter for comment and satisfaction. But the South had a long way to go; 
and it did not go far enough or fast enough in view of the size and density of 
the population on the land. Thus in 1910 the Southeast was still the least indus- 
trialized of all the regions in the sense that the manufacturing-services com- 
ponent of its labor force remained the lowest in the nation. 

It was only in forest and logging activity that the region as a whole began 
to achieve a superior position relative to the rest of the nation. Florida, and 
later West Virginia, developed heavy concentrations of loggers and, reinforced 
by developments in Arkansas and Louisiana, brought lumbering in the South- 
east rapidly up above the national level of forest activities. After 1890 the influx 
of logging activity stemmed directly from the exhaustion of the pinelands in the 
northern Great Lakes states. 2 By 1910 the proportion of loggers and other forest 
workers in the Southeast's labor force had reached a peak not to be surpassed 
much before 1940. 

In absolute terms, miners outnumbered woodsmen and loggers in the South- 
east during the 1870-1910 period. The proportion of miners in the regional 
labor force, however, was less than that in the national labor force. But West 
Virginia was among the great mining states. Even before 1870 it exceeded the 
national average in mining employment, and by 1910 only two states — Arizona 
and Nevada — had a greater proportion of their labor forces occupied in mining 
than West Virginia. By 1900 Alabama also exceeded the national average. Ken- 
tucky, Tennessee, and Virginia, on the other hand, failed to achieve the national 
level of mining employment over these years. 

The absolute size of the Southeastern fisheries was small throughout the 
period under review. But after 1880 the region had a much greater relative 
concentration of fishing employment than the nation as a whole. Virginia, 
Florida, and Louisiana were among the leading fishing states; before 1890 they 
were joined by North Carolina and shortly thereafter by Arkansas, which had 

2 F. W. Kollmeyer, "Northern Pine Lumbermen: A Study in Origins and Migration," 
Journal of Economic History, Vol. 16 (1956), pp. 529-38. 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 111 

developed important river fisheries. 

Throughout the 40-year period Mississippi had the greatest concentration of 
agricultural workers (85% in 1870 and 76% in 1910) and the lowest concen- 
tration of manufactures (3% in 1870 and only 7.5% in 1910). Elsewhere in 
the Southeast the structure of the state labor forces changed rapidly from decade 
to decade. At the beginning of the period in 1870 West Virginia had the greatest 
concentration of manufactures (14%) and Louisiana was the least agricultural 
of the Southeastern states (61% in agriculture). By 1890, Virginia was most 
heavily concentrated in manufactures (15.26%) and also the least in agricul- 
ture (51%). By 1910, however, Florida had developed the greatest concentra- 
tion in manufactures (22.8%) and West Virginia had the smallest proportion 
of its labor force in agriculture (35.6%). In absolute terms, Virginia had the 
largest manufacturing contingent in 1910, almost 162,000 workers. 

The Southeast had started late down the road to industrialism; there was a 
chronic scarcity of skilled labor and liquid capital. The region had begun first- 
stage processing of its bulkier materials, but apart from the furniture and fix- 
tures industry in North Carolina and a few textile centers, it was as yet rare for 
the Southeast to produce finished products for the rest of the nation. Recent 
students of Southern industry differ as to the precise reasons for this prolonged 
backwardness. But they are in general agreement that the region was still a 
tributary raw-material economy dominated by absentee owners and suffering the 
penalties attendant on "backwardness": low wages, lack of opportunity, and 
general poverty. 3 



Industrial Structure of the Labor Force in the Plains Region, 1870-1910 

In some respects the Plains region resembles the Southeast more than any 
other part of the country in its industrial character (Table 62). It differs, how- 
ever, in that it supports a much smaller population and is culturally more akin to 
the western Great Lakes states than to the Southeast. It is primarily a region 
of agricultural staples, although Missouri, Kansas, and Minnesota have or have 
had major deposits of important minerals such as coal, oil, and iron ore. The 
region's slight forest cover was rapidly depleted in the course of settlement, and 
its fisheries are negligible. Manufacturing has been confined mainly to proc- 
essing activities in a relatively few urban-industrial centers. Service employ- 
ment has been at a much higher level than in the Southeast; in most years, close 
to the national average. The farm lands of Iowa and adjacent states are among 
the best in the country, but beyond the Missouri Valley the Plains merge into the 
"Great American Desert" and agriculture becomes a hazardous business for the 
small cultivator. 

3 H. L. Herring, Southern Industry and Regional Development (Chapel Hill: University 
of North Carolina Press, 1940), pp. 31-5; C. Vann Woodward, Origins of the New South, 
1877-1913 (Baton Rouge: Louisiana State University Press, 1951), pp. 291-320; W. H. 
Joubert, Southern Freight Rates in Transition (Gainesville: University of Florida Press, 
1949), pp. 1-134; and C. B. Hoover and B. U. Ratchford, Economic Resources and Policies 
of the South (New York: Macmillan, 1951), pp. 65-88. 



178 / Regional Economic Development: 1870-1950 

Table 62. Distribution by industry of Plains region labor force, 1870-1910 













(per 


cent of total) 












Manu- 




Year 


Agriculture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


61.63 


0.50 


0.26 


0.03 


14.91 


22.67 


1880 


61.35 


0.95 


0.26 


0.02 


14.34 


23.08 


1890 


51.57 


1.29 


0.33 


0.04 


16.42 


30.35 


1900 


49.87 


1.69 


0.34 


0.06 


16.75 


31.29 


1910 


40.85 


1.82 


0.29 


0.04 


19.99 


37.01 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



As late as the 1880's the western reaches of the Plains were still undergoing 
settlement and, while the agricultural sector of the labor force in Minnesota, 
Iowa, and Missouri was already in decline, that of the Dakotas, Nebraska, and 
Kansas were still growing. By 1890 agricultural employment was shrinking rela- 
tively in every part of the Plains except the Dakotas and continued to decline 
thereafter. By 1910 it made up less than half of total employment in every state 
except North and South Dakota, and in Minnesota and Missouri it was barely 
one-third of the total. 

Mining surged forward rapidly in this region in the 1870's owing to the fervid 
search for precious metals in the Dakotas. In 1880 mining employed 7.3% of 
the territorial labor force, but by 1890 the proportion of South Dakota's labor 
force employed in mining had declined to 2.3%. Less spectacular digging had 
raised Missouri's mining contingent over these years to 1.8% of its total labor 
force, by 1890 and that of Kansas and Iowa to almost 1.5% each. Before the 
close of the century mining employment had risen to 2.3% of total employment 
in Missouri and 2.2% in Kansas; it had declined to 2.1% of the total in South 
Dakota. By 1910 the focus of concentration in mining in the region had turned 
northward to the iron fields of Minnesota (2.3%), the oil and coal fields of 
Kansas (2.2%), and the coal fields of Missouri (2.08%). Only North Dakota 
and Nebraska, in fact, then had a negligible proportion of miners in their labor 
forces. 

Forest employment accounted for from 1% to 1.2% of total employment in 
Minnesota. Elsewhere in the Plains the forestry sector made little contribution 
to employment, although for a brief period around 1870 the small labor force 
of the Dakotas contained a logging contingent of about 2.28%, probably as an 
adjunct to mining activity. 

The early promise of the Plains region was based on its rich agricultural lands 
and the occasional outcrops of mineral wealth. Yet as early as 1870 its indus- 
trial component — manufactures and services — amounted to more than 37% of 
the regional labor force and by 1910 to 57%. It was in services, however, rather 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 179 

than manufactures that the greatest advances were made and where parity with 
the nation was achieved after the slump of the seventies. Missouri and Min- 
nesota were always the region's leading manufacturing states. After 1890 they 
had from 19% to 24% of their labor forces in manufacturing, whereas the 
Dakotas did not have 10% of their workers in manufactures before the end of 
the century. Minnesota and Missouri were likewise the leading centers of service 
employment, though the extension of railroads and expansion of trade in 
Nebraska, Kansas, and Iowa slowly increased the concern of these states with 
services as well. The Plains region, therefore, like the Southeast was primarily 
a resource region during this period, with little manufacturing except in the 
larger metropolitan centers of Missouri and Minnesota. But, unlike the South- 
east, its population densities were relatively low; this permitted it to achieve a 
considerably higher level of personal income per capita than the Southeast. 



Industrial Structure of the Labor Force in the Southwest, 1870-1910 

The industrial structure of the labor force in the Southwest, shown in Table 
63, characterizes this region also as essentially a resource region. The growth of 
its manufactures before 1910 was at least as unimpressive as that of the South- 
east. By 1910, however, its proportion of employment in manufactures and 
services combined was running ahead of the Southeast's by almost 2 percentage 



Table 63. Distribution by industry of Southwest region labor force, 1870-1910 













(per 


cent of total) 












Manu- 




Year 


Agriculture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


73.67 


0.31 


0.16 


0.02 


7.04 


18.80 


1880 


72.33 


1.29 


0.27 


0.02 


6.41 


19.68 


1890 


63.94 


1.14 


0.30 


0.06 


9.11 


25.45 


1900 


66.71 


1.65 


0.38 


0.05 


8.35 


22.86 


1910 


57.67 


1.77 


0.36 


0.05 


12.37 


27.78 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



points. In area, the Southwest is one of the largest of all the eight regions, 
covering nearly 20% of the continental United States, but it has one of the 
lowest population densities. Hence it represents a very different type of resource 
economy from that of either the Southeast or the Plains. 

Until well into the present century the Southwest, apart from Texas, was a 
political dependency of the federal government. Oklahoma achieved statehood 
in 1907, Arizona and New Mexico only in 1912. The census provides very little 
data on Oklahoma before 1890, since it was mostly a separate Indian territory. 



180 / Regional Economic Development: 1870-1950 

The region as a whole was overwhelmingly agricultural. Yet as early as 1870, 
because of work on the railroads, the Arizona Territory had more than twice 
as many workers in transportation, trade, and manufactures as in agriculture. 
Mining developments were also under way. Although the preponderance of the 
manufacturing-services component in Arizona endured into the nineties, miners 
comprised more than one-fourth of the territorial labor force by 1880, and as 
late as 1910 mining provided nearly 18% of all employment. Meanwhile, devel- 
opments in Texas and the later opening of farm lands in Oklahoma had begun 
to affect the industrial structure of the region as a whole. By the early nineties, 
agricultural workers comprised about two-thirds of the labor force in these two 
areas and more than half the labor force of the neighboring territory of New 
Mexico. The Southwest, in fact, like the Mountain and Far West regions, was 
increasingly agricultural down into the opening years of this century. Arizona's 
labor force also contained the largest proportion of workers in forestry and 
logging, although Texas had the largest numbers involved, Texas likewise 
contained a vast majority of the region's fishery workers. 

The contribution of the Southwest to the growth of the national economy in 
these years was made primarily through its mineral resources — at first copper, 
and later oil. Only the Mountain and Far West regions ever achieved greater 
concentrations of mine workers than the Southwest; and only individual states 
like Pennsylvania and West Virginia developed comparable concentrations of 
miners east of the Mississippi (though actual numbers of mine workers in the 
great coal-producing states of the East, of course, greatly exceeded the con- 
centrations of the West). Arizona and, to a lesser extent, New Mexico, were the 
chief mining territories of the Southwest, but after 1900 both Texas and Okla- 
homa increased their contribution to mineral wealth through the medium of oil 
development. The economic development of Texas was also a major factor in 
the growth of the entire Southwest in both the argricultural and the industrial 
sectors. 



Industrial Structure of the Mountain Region s Labor Force between 1870-1910 

The structure of the Mountain region's labor force between 1870 and 1910 
is shown in Table 64. The proportion of the regional labor force in agriculture 
was below the national average throughout these years. Unlike the Southwest 
and the Far West, the Mountain region was not settled primarily by agricul- 
turalists. Mining and services provided the principal means of subsistence until 
the early seventies, and in both sectors the region was an "exporter" in the sense 
that even its services were in large part used in facilitating the development of 
regions around it. 

In 1870 miners outnumbered the combined totals of farm and service workers 
in Idaho and Montana, while in Colorado they almost equalled the numbers in 
manufactures. The Mormon commonwealth of Utah was at once the largest 
agricultural and the largest manufacturing territory in the region; the thinly 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 181 

Table 64. Distribution by industry of Mountain region labor force, 1870-1910 

(per cent of total) 













Manu- 




Year 


Agriculture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


33.59 


26.54 


0.84 


0.02 


12.94 


26.07 


1880 


24.00 


26.37 


0.97 


0.02 


18.02 


30.62 


1890 


26.53 


12.71 


1.18 


0.02 


19.99 


39.57 


1900 


30.12 


14.72 


0.73 


0.01 


18.68 


35.74 


1910 


29.93 


8.47 


0.67 


0.02 


20.48 


40.43 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 

populated Wyoming territory had a striking concentration in tertiary employ- 
ment, chiefly transportation workers. 

By 1890 Colorado had acquired the greatest concentration of manufactures, 
and Idaho had the greatest concentration in agriculture. All states had large 
volumes of service employment, ranging from 27% in Idaho to 42.4% in 
Colorado. Utah had the smallest stake in mining (6.5%); in its neighboring 
states, mining activities ranged upward from 11.7% in Wyoming to more than 
17% in Idaho and Montana. 

By 1910, all states in the region except Idaho (32.8%) had about 40% of 
their working force in services; all except Idaho, once more, had from one-third 
to one-fourth of their labor forces in agriculture; Idaho had 42%. The greatest 
concentrations in manufactures were in Utah (23%) and Colorado (22%); 
Wyoming had the least (15%). The largest mining contingents were in Mon- 
tana and Wyoming (just over 10% each) ; the smallest was in Idaho (about 
5%). Idaho and Montana had the largest proportions of forest workers; in 
Idaho they constituted more than 2% of the labor force in 1910. These two 
states had, in fact, dominated the region's small forest economy since the decline 
of Wyoming a decade or so before. 

Thus the semi-arid Mountain region was another of the great resource regions 
of the United States in the 1870-1910 period. The development of its material 
wealth during this period indicates that it contributed much to the growth of 
the industrial economy in other parts of the country. As a consequence, its own 
modest manufactures largely took the form of first-stage processing in prepara- 
tion for later processing and finishing in other regions, e.g., copper smelting. 
Much of the necessary fuel supply could be found locally in one or other of the 
Mountain states, but before 1910 little effort was made to exploit the region's 
known oil potential. 

Like the Southwest, the Mountain region was also an important "corridor" 
between the East and Far West. For this reason it developed important trans- 
portation and communications services which helped bind the national economy 
together. Yet nature had placed both the Mountain and Southwestern regions 



182 / Regional Economic Development: 1870-1950 

in the heart of an arid and semi-arid expanse. That the desert might be turned 
into a garden had been proved long before 1870 by the Mormon pioneers in 
Utah; what they and their successors had not proved, however, was the capacity 
of the region to support vastly increased numbers of people at higher levels of 
living. 



Industrial Structure of the Labor Force in the Far West,, 1870-1910 

The industrial structure of the Far West's labor force is shown in Table 65. 
The enormous contribution in resources which this region made to American 



Table 65. Distribution by industry of Far West region labor force, 1870-1910 













(per 


cent of total) 












Manu- 




Year 


Agriculture 


Mining 


Forestry 


Fisheries 


facturing 


Services 


1870 


25.73 


18.03 


1.63 


0.41 


18.20 


36.00 


1880 


28.81 


11.46 


1.65 


1.35 


20.45 


36.28 


1890 


28.85 


4.44 


2.30 


0.54 


22.55 


41.32 


1900 


27.65 


5.27 


2.17 


0.76 


22.06 


42.09 


1910 


20.10 


2.84 


2.04 


0.47 


26.95 


47.60 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



growth during the 1870-1910 period can be judged from the fact that it had 
higher concentrations than the national average in three of the four resource 
sectors: mining, forestry, and fisheries. 

But the strength of the Far West during the period here under review was by 
no means confined to the resource sectors of the economy. This can be seen by 
comparing the combined manufacturing and services component of its labor 
force with that of each of the three most industrialized regions — the New Eng- 
land, Middle Atlantic, and Great Lakes regions (Table 66) . Though less indus- 
trialized, in this sense, than either the New England or the Middle Atlantic re- 
gions, the Far West was nevertheless more industrialized than the dynamic Great 
Lakes region. Its strength, however, was in services rather than manufactures. 

The remoteness of the Far West from the industrial center of the Northeast 
meant that it had to develop certain manufactures which it could not secure 
from the rest of the economy except at almost prohibitive cost. By 1870, there- 
fore, it had achieved the greatest concentration of manufacturing-service activ- 
ities outside the Northeastern industrial belt. The fact of distance, measured in 
terms of costs of the long haul to the Coast by land or sea, had the effect of 
placing a "protective tariff" around some forms of Far Western manufactures. 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 183 

Thus the economy of the region developed somewhat as an empire apart from 
the rest of the nation. 

California was more industrialized than any of its neighboring states, or 
indeed than the United States as a whole, during this period. 4 But by 1890 
Washington had emerged as a second industrial state, with sizable shares of 
employment in both manufactures and services; and after 1900 Oregon's pro- 
portion of manufacturing-services workers also exceeded the national average. 
By 1910 the three Pacific Coast states were among the nation's leading industrial 
areas. Nevada's small labor force contained a larger proportion of miners than 
the labor force of any other state in the nation (21.9%). 



Table 66. Combined manufacturing-services components of labor force of Far 
West, New England, Middle Atlantic, and Great Lakes regions, 1870, 1890, 
and 1910 











(per 


cent of total) 


Year 


Far West 


New England 


Middle Atlantic 




Great Lakes 


1870 
1890 
1910 


54.20 
63.87 
74.55 


72.07 
82.06 
89.30 


68.54 
77.98 
85.36 




42.79 
57.59 
71.82 



Source: Appendix Tables Al, A6, and A7. 



At various times during the 1870-1910 period all states in the region made 
large contributions to the nation's lumber supply. By the end of the nineteenth 
century, however, the preponderance of forest activities had passed to Washing- 
ton and Oregon. During the present century, in fact, forest industry products 
have regularly yielded a large proportion of the income of these two states. The 
same two states usually had the region's largest concentrations of fishing em- 
ployment (including the lucrative salmon fisheries) and at all times had large 
numbers employed in various specialized branches of agriculture. By 1910, 
Oregon had the largest proportion of farm workers of any state in the region. 
Needless to add, California was also one of the great fishing, forest, and farming 
states, but its rapidly growing shares of employment in manufactures and serv- 
ices tended to reduce the relative significance of its numbers in the resource 
sector even before the turn of the century. 

In celebrating the great material endowment of the Far West, it is well to 
conclude by stressing at least one major disadvantage. With practically no 
commercial supply of coal, the Far Western states were for long without an 
adequate or assured fuel base for their growing industries and communications 

"See R. G. Cleland and Osgood Hardy, March of Industry in J. R. McCarthy, ed., Cali- 
fornia (Los Angeles, Calif.: Powell Publishing Co., 1929), on early industrial development 
in California. 



184 / Regional Economic Development: 1870-1950 

systems. Before the close of the century, however, California's oil potential was 
already being developed; production soared from 4,000,000 barrels in 1900 
to 55,000,000 barrels in 1910. Natural gas also soon came to be a major source 
of domestic heat in the region's burgeoning urban-industrial centers. Oregon 
and Washington also lacked ample coal for industrial purposes and the defi- 
ciency had to be made up by importing coal from British Columbia and oil from 
California. The abundant rainfall and strong river currents also combined to 
make the Far West (especially the Northwest) a great potential producer of 
hydro-electric power. The development of this resource, however, lies outside 
the period which ended in 1910. 



Income Changes, 1880-1920 

Changes in total personal income and per capita personal income provide a 
rough guide to changes in the regional volume of activities and levels of living; 
and these, in turn, are related to the structural changes in population and labor 
force which we have described. We shall now turn to a consideration of some of 
these relationships during the period 1880-1920. 5 

Data on the regional distribution of total income over these years are shown 
in Table 67. The over-all trend in the regions clearly reflects the growth of total 
personal income in the nation as a whole. The rate of growth was, in general, 
greater between 1900 and 1920 than between 1880 and 1900. Among the more 
populated regions, the agricultural Plains probably experienced the fastest 
relative growth in income before 1900, surpassing both New England and the 
Southeast, in absolute terms, by 1900. Thereafter, however, it slowed consid- 
erably, and by 1920 its share in the total national income was somewhat smaller 
than in 1880. The Southeast, like the Plains, had a smaller share in 1920 than 
in 1880; its decline, however, occurred entirely within the earlier half of the 
period. New England's relative share declined throughout these years. Among 
the newer Western regions, the Far West and Southwest increased their relative 
shares both between 1880 and 1900 and between 1900 and 1920; their gains 
were especially marked after 1900. The Mountain region, like the Plains, rose 
between 1880 and 1900 and dropped back after 1900. Its decline, however, in 
the second of these periods was by no means as great as that of the Plains, and 
its relative share in 1920, though lower than in 1900, was higher than in 1880. 

5 Unfortunately, for the first half of our period, the years for which income data by states 
are available do not coincide exactly with the years for which population and labor force 
data can be obtained. Thus, figures of state income payments, devised by Richard Easterlin, 
begin with 1880 and proceed at two-decade intervals until 1920; thereafter figures become 
available by decades and after 1929, in fact, on an annual basis in the Commerce series. 

See E. S. Lee, A. R. Miller, C. P. Brainerd, and R. A. Easterlin, Population Redistribu- 
tion and Economic Growth, United States, 1870-1950 (Philadelphia: American Philosophical 
Society, 1957), pp. 703-45; C. F. Schwartz and R. E. Graham, Jr., Personal Income by 
States since 1929, a supplement to the Survey of Current Business, Office of Business 
Economics, U. S. Department of Commerce (Washington: Government Printing Office, 
1956). 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 185 

Table 67. Regional distribution of total personal income in the United States, 
1880, 1900, and 1920 

(U. S. income in millions) 



Region 


1880 


1900 


1920* 


United States 


$8,743 


$15,391 


$69,276 




100% 


100% 


100% 


New England 


11.30 


9.86 


8.75 


Middle Atlantic 


32.60 


30.80 


30.20 


Great Lakes 


22.84 


22.45 


22.21 


Southeast 


13.66 


12.02 


13.02 


Plains 


11.07 


13.27 


10.27 


Southwest 


2.11 


3.76 


5.69 


Mountain 


1.43 


2.53 


2.47 


Far West 


4.99 


5.31 


7.39 



* 1919-21 average. 
Source: Lee, Miller, Brainerd and Easterlin, Table Y-l. (See note to Table 8.) 



The Middle Atlantic and Great Lakes regions grew as the nation grew 
throughout the entire period. Their combined share of the nation's total income 
in 1880 was 55%; and in 1920, after four decades of rapid growth and redistri- 
bution of economic activity, it still exceeded 52%. In absolute terms, these two 
regions enjoyed the largest total personal income throughout the 40-year period. 

Table 68, which shows the principal net upward and downward shifts among 
the states during this period, highlights the more striking changes. Almost half 
of the net upward shift (relative growth) occurred in four states in the South- 



Table 68. Principal net shifts in total personal income among states, 1880-1920 





Percentage net 




Percentage 


net 






upward 


shift 




downward 


shift 




Texas 






16.15 


Pennsylvania 




13.20 


Oklahoma 






13.02 


Iowa 




10.37 


Washington 






11.48 


Missouri 




9.11 


California 






9.34 


Indiana 




8.25 


Dakotas 






5.60 


Massachusetts 




8.20 


Michigan 






4.61 


New York 




7.97 


W. Virginia 






4.00 


Kentucky 




5.43 


Illinois 






3.89 


Ohio 




5.38 


New Jersey 






3.68 


Maine 




3.82 


Minnesota 






3.61 


Louisiana 




3.35 



Source: Same as Table 67. 



186 / Regional Economic Development: 1870-1950 

west and Far West— 29.17% of it in Texas and Oklahoma and 20.82% in 
California and Washington. These same four states also accounted for almost 
43% of the total population shifts during the 1870-1910 period. Altogether, five 
of the ten states with the greatest relative income growth between 1880 and 1920 
were also among the ten with the greatest relative population growth between 
1870 and 1910. And seven of the ten states with the greatest relative declines in 
total personal income between 1880 and 1920 were among the ten with the 
greatest relative declines in population in the 1870-1910 period. 

The weakness of personal income growth in the Plains region after 1900 
was somewhat localized, as the figures in Table 68 show; it stemmed from the 
relative decline in the key states of Iowa and Missouri, which together contrib- 
uted almost 20% of the total net downward shift in personal income among 
the states during the 1880-1920 period. Similarly, in the Great Lakes region, the 
relative strength of Michigan and Illinois did not entirely offset the relative 
downward shifts in Indiana and Ohio (Figure 31). In the Southeast, also, there 




Absolute upward shift 
l'--''v>'l Relative downward shift 



Figure 31. Net Shift in Total Personal Income, 1880-1920. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 29.2. Total net shift as % of 1880 personal income: 90.1. 



were a number of income growth states between 1880 and 1920 — West Virginia, 
Florida, North Carolina, and Arkansas in that order — but their combined up- 
ward shift of 10.62% did not offset the relative declines registered by the other 
states in that region. Relative decline in income was in some respects even more 
pronounced in the two seaboard regions of the Northeast. New Jersey was the 
only state among ten which did not experience a net downward shift in personal 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 187 

income; Pennsylvania, Massachusetts, and New York, in fact, contributed more 
than 29% of the total net downward shifts in income among the states during 
this period. 

The relation between the relative growth in total population and in total 
income within a state or region, as shown in the figures for personal income 
per capita, provides a valuable, if rough, measure of changes in the levels of 
living. Table 69 shows how the regions ranked in terms of real income per 
capita as compared with the national average during the 1880-1920 period. The 
Far West, Mountain, New England, Middle Atlantic, and Great Lakes regions 
exceeded the national average throughout these years. The Plains, Southeast, 
and Southwest regions consistently fell below. 

In the above-average group, the greatest relative gains were made by the 
urban-industrial regions of the Northeasten seaboard, and more especially by 
the heavily populated Middle Atlantic region. By 1900 the increasingly urban- 
industrialized Great Lakes region was also making rapid gains in personal 
income per capita, but the sparsely populated rural-mining Mountain region 
was already in absolute decline. The Far West was the only one of the newer 
Western regions that maintained its position in the above-average group; and 
its greatest growth occurred after 1900, by which time it too had developed a 
labor force structure resembling the urban-industrial pattern of the Northeast. 

In the below-average group, the Plains region came closest to the national 
average. Around 1900 its per capita income reached 97% of the average for 
the nation as a whole; by 1920, however, it had fallen to less than 90%. The 
Southwest and Southeast, in contrast to the Plains, did somewhat better after 
1900 than before. The Southwest, to be sure, was closing the gap between its 
per capita income level and the national level throughout the period 1880-1920, 
but the Southeast began to converge on the national average only around the 



Table 69. Rank order of regions by personal income per capita,* 1880, 1900, 
and 1920 











(1929 dollars) 


1880 




1900 




1920 




Far West 


638.2 


Far West 


674.6 


Far West 


780.6 


Mountain 


501.5 


Mountain 


601.0 


Middle Atlantic 


772.5 


New England 


425.0 


Middle Atlantic 


570.2 


New England 


719.3 


Middle Atlantic 


422.8 


New England 


551.7 


Great Lakes 


623.5 


Great Lakes 


307.4 


Great Lakes 


439.6 


Mountain 


590.7 


United States 


302.1 


United States 


414.0 


United States 


578.0 


Plains 


271.2 


Plains 


401.5 


Plains 


501.0 


Southwest 


182.7 


Southwest 


283.2 


Southwest 


466.2 


Southeast 


150.6 


Southeast 


197.7 


Southeast 


326.2 



*Weighted averages of state incomes per capita. 
Source : Unpublished data by R. A. Easterlin. 



188 / Regional Economic Development: 1870-1950 

turn of the century. In spite of impressive growth in total personal income, the 
depressing effect of population in the Southeast held the per capita figure far 
below the national average in both relative and absolute terms. 

We turn finally to a consideration of certain relationships between demo- 
graphic patterns, occupational structure of the labor force, and levels of per 
capita personal income as they have emerged in our survey of economic changes. 
The associations between urban-industrial structure and levels of living in 1920 
are set out in Table 70. There is evident the close relationship between the 
degree of urbanization and the industrialization of the labor force and, to a 
lesser extent, between the degree of urbanization and the quantitative decline in 
agricultural employment. The Far West, however, deviates somewhat here from 
the rest of the country. There is also evident a close association between urban- 
industrial structure and high levels of personal income per capita and, by the 
same token, between rural-agricultural structure and relatively lower levels of 
per capita personal income. There is, however, no apparent association between 
the non-agricultural resource sector (mining, fisheries, forestry and logging) 
and urban-industrial structure or personal income per capita. But, read in con- 
junction with certain other occupational-structural and demographic features, 
the data on the non-agricultural resource sector may throw some light on the 
income performance of the Far Western and Mountain regions. 

The Far West is of special interest with regard to these relationships. As it 
was only the third most urban-industrialized region in the period under review, 
it might be expected to rank below the top in average personal income. But it 



Table 70. Rank order of regions by personal income per capita, urbanization 
of population, and labor force industry components, 1920 





Personal 




Labor 


force components 
















per capita 


proportion 




% Non- 






(1929 


of 




agric* 


% Mfg. & 


Region 


dollars) 


population 


% Agric. 


resource 


Services 


United States 


$578.0 


51.2% 


25.63 


3.27 


71.10 


Far West 


780.6 


62.6(3) 


18.33(6) 


3.99(2) 


77.68(4) 


Middle Atlantic 


772.5 


74.7(2) 


7.38(7) 


3.69(4) 


88.93(2) 


New England 


719.3 


75.9(1) 


6.93(8) 


1.12(8) 


91.95(1) 


Great Lakes 


623.5 


60.8(4) 


18.72(5) 


3.00(6) 


78.28(3) 


Mountain 


590.7 


39.7(5) 


32.85(4) 


7.73(1) 


59.42(6) 


Plains 


501.0 


37.7(6) 


36.35(3) 


2.08(7) 


61.57(5) 


Southwest 


466.2 


30.3(7) 


44.88(2) 


3.81 (3) 


51.31(7) 


Southeast 


326.2 


23.9(8) 


49.50(1) 


3.59(5) 


46.91(8) 



Figures in italics are above the national average. 
* Forestry and logging, mining, and fisheries. 
Rank order in parentheses. 

Source: Unpublished data by R. A. Easterlin; 17th U. S. Census, 1950, see supra, Table 4, 
note; and Appendix Tables A1-A7. 



Changing Regional Structure of U. S. Economy, 1870-1910 (3) / 189 

ranks first. An explanation of this, we believe, is to be found in a combination 
of historical and geographical circumstances peculiar to this region. Economic 
development of the Far West got under way during the third quarter of the last 
century after the initial discovery of precious metals, but it did not gather great 
momentum before the completion of railroad connections with the East around 
1870. At this time, the Far West's share of total population and its level of popu- 
lation density were lower than those of any other region except the Mountain; 
these same relative conditions still obtained during the second decade of the 
twentieth century. Thus crowding on land, which has been an income-depressing 
factor in other regions, did not occur in the Far West. The region was also rich 
in minerals, forest, and agricultural land ("desert" was widespread only in the 
sparsely populated state of Nevada and in parts of southern California). This 
fortunate combination of low population and abundance of resources lifted the 
Far West, and to a lesser extent the Mountain region, to the highest per capita 
income levels anywhere in the United States between 1880 and the end of the 
century. 6 

The urban-industrial regions of the Northeast, and especially the Middle 
Atlantic region, had drawn upon their own and other regions' resources to raise 
their much larger populations to average levels of personal income not so very 
far below those obtaining in the Mountain or Far West regions. Before 1910, in 
fact, the Middle Atlantic and New England regions had surpassed the per capita 
income levels obtaining in the Mountain region, and before the close of the next 
decade the Middle Atlantic region was challenging the Far West for first place. 
By 1920 again, the rapidly industrializing Great Lakes region had also surpassed 
the income levels of the Mountain region and had climbed into fourth place 
behind the Far West, Middle Atlantic, and New England regions. 

The relative decline of the Mountain region and the accelerated growth of the 
Far West after 1900 are reflected in the occupational structure of their respec- 
tive labor forces (see Table 71). Between 1900 and 1920 the Far West main- 
tained its development along urban-industrial lines, whereas the Mountain 
region experienced only a modest increase in the industrial-services sector, 
and that increase was primarily in services rather than in manufactures. After 
oil was developed in California and transportation had been facilitated by the 
completion of the isthmian canal, the Far West could realize its industrial 
potential; but the Mountain region remained relatively inaccessible and was 
handicapped by the partial exhaustion of some important mineral resources. Yet, 
compared with other predominantly non-industrial regions (Table 70), the 
Mountain region continued to enjoy relatively high levels of per capita income 
at least until 1920. Its labor force structure was clearly not following the typical 
pattern of the older urban-industrial regions, but neither was it settling down 
into the pattern of the older rural-agricultural regions. As long as this region 
would maintain its rich non-agricultural resource sector and hold its population 
increase in bounds, it might hope to go on enjoying higher average levels of 

6 No doubt, the high average levels of personal income obtaining in Far West and Moun- 
tain states also reflect the high prices which prevailed in these relatively inaccessible parts 
of the country. 



190 / Regional Economic Development: 1870-1950 

Table 71. Resource and non-resource components of the labor force in the 
Mountain and Far West regions, 1900 and 1920 

(per cent of total) 





1900 




1920 




Region 


Resource 


Other 


Resource 


Other 


Mountain 
Far West 


45.58 
35.85 


54.42 
64.15 


40.58 
22.32 


59.42 
77.68 



Source: Appendix Tables A1-A7. 



personal income than the more heavily populated, rural-agricultural regions, 
such as the Plains, Southwest, and Southeast. 

These suggestive associations between the industrial structure of a region and 
its level of per capita income raise many questions concerning the human, 
material, and institutional forces that underlie the development of a given 
industrial structure. Certain of these questions are probed in some detail in 
later sections of this book. Here it will suffice to say that, with one exception, 
no region that was significantly below the national average in urban-industrial- 
ization was able to exceed the national average in per capita income level. The 
exception was the Mountain region, much the most sparsely populated part of 
the country and a region, moreover, in which the non-agricultural resource com- 
ponent of the labor force as late as 1920 was more than twice the relative size of 
the non-agricultural resource component of any other region. Between 1900 
and 1920, however, the Mountain region was the least buoyant of the eight 
regions. As the relative size of its agricultural labor force grew, its level of per 
capita personal income fell. It had a larger share of the nation's population in 
1920 than in 1900, but it was the only region that had a relatively larger agri- 
cultural labor force in 1920 than in 1900 and it was the only region to experi- 
ence a falling average level of per capita income. After 1910, the relative size of 
its manufacturing labor force also declined. Thus, between 1910 and 1920, the 
Mountain region was unique in that its manufacturing was already in relative 
decline while its agriculture continued in relative growth. 



14 Nationalizing the American 

economy, 1870-1910: 

THE PLAY OF TECHNOLOGY ON RESOURCES 



We have traced the growth and redistribution of population, productive activ- 
ities, and personal income that attended the evolution of the continental economy 
in the United States between 1870 and 1910. We shall now consider the role of 
natural resources, and especially the resources of the West, in the general 
economic history of these years. 

For practical purposes, the joining of the Union and Central Pacific railroads 
in Utah on May 10, 1869, may be said to symbolize the beginnings of a truly 
nation-wide economy. 1 Subsequent preoccupation with the vast resource poten- 
tial made accessible by the completion of the railroad network did not mean, of 
course, that the country severed its close ties with the Atlantic economy. On the 
contrary, the movements of population, capital, and goods across the Atlantic 
reached unprecedented levels during the late nineteenth century. The pace and 
character of American growth, however, were no longer determined so largely by 
the size and kinds of overseas demand. Exports rose in volume throughout the 
rest of the century, but they were an increasingly smaller portion of total Ameri- 
can output. More significant, though the United States remained the world's 
largest debtor, its economic development was now less dependent upon con- 
tinuing imports of European capital. Billions of dollars flowed into the economy 
concurrently with millions of immigrants, but indigenous capital formation 
seldom comprised less than 95% of investment increase in any one year. Indus- 
trial Europe, meanwhile, became more and more dependent on supplies of 
American farm staples. Technical developments over these years enabled the 
United States to meet its mounting foreign obligations with a declining 
proportion of its greatly enlarged output. 

1 A transcontinental telegraph was opened in October 1861. Before that time it required 
at least three weeks for mails to reach San Francisco from the Mississippi Valley by the 
Overland Mail or from the Atlantic seaboard via the Pacific Mail and Panama route. The 
"Ponv Express," inaugurated in 1860, could carry urgent communications in about eight 
days, but heavy duty freight movements from coast to coast were impossible except by long 
ocean passage. 

/ 191 



192 / Regional Economic Development: 1870—1950 

The forces of population growth and of technical and economic progress 
that had been operative in the crucial political-economic decision of the sixties 
were now free to shape the continental framework of the nation's economic 
growth. 2 On average, net national product per capita increased over 10% per 
decade over the years 1870-1910. After 1890, when the so-called "frontier of 
settlement" finally disappeared, the advertising of numerous products was under- 
taken on a nation-wide scale, and by the early years of the 20th century the 
domestic market embraced all the population. 

Some historians have suggested that in the years after 1860 the urban-indus- 
trial Northeast imposed its way of life upon a defenseless rural America. But 
this essentially "political" point of view exaggerates the consensus among 
Northeastern manufacturers in regard to such issues as tariff and monetary 
policy; the interests of New England textiles and Pennsylvania iron in these 
issues, for example, were by no means identical. Moreover, a national banking 
system, land grants to transcontinental, and the easing of state incorporation 
laws were not so much examples of "sectional imperialism" as of belated and, 
on the whole, successful efforts to create institutions by means of which pro- 
ductive activities could be organized on a continental scale. If sectional conflicts 
persisted, and for a while at least they were aggravated by agrarian politics, 
they were nevertheless a waning influence, complicating administration and 
policy, but always giving way before the surge of industrial growth. By the 
early years of this century modern urban-industrial America had been achieved : 
a powerful continental economy made up of a complex of differentiated regions 
that could offer families in every part of the country a more abundant material 
life. 

This enormous transformation was not, of course, effected merely by the 
impersonal forces of industrial revolution. It was also the lifework of millions 
of men and women whose motivations and mores were moulded in turn by the 
social process they helped forward. No period of American history has produced 
a more dramatic human spectacle than the decades following the Civil War; at 
no other time perhaps did "economic" man so completely dominate the Ameri- 
can scene. But the heroes of this "age of enterprise" were no longer provincial 
merchants gathering in a material surplus for trade beyond the seas; they were 
the promoters of continental empires in transportation, finance, natural 
resources, and manufactures. The Astors were no longer an exceptional breed; 

2 At the turn of the century the Final Report of the U. S. Industrial Commission (Wash- 
ington: Government Printing Office, 1902), Vol. 19, p. 1, concluded that the country had 
been occupied and that remaining "areas of unsettled country" were chiefly (1) moun- 
tainous regions unsuited to agriculture or inaccessible to markets; (2) swamp and lake 
regions; (3) timber lands; and (4) arid lands. In fact, more land was taken up under 
terms of the Homestead Act and its successors after 1890 than ever before. As late as 1913 
the Secretary of Agriculture reported that less than 60% of arable land was under cultiva- 
tion and that, of this, not more than 12% was yielding "reasonable full returns." See D. F. 
Houston, Eight Years with Wilsons Cabinet (New York: Doubleday Page and Co., 1926), 
p. 200; L. P. Jorgenson, "Agricultural Expansion into the Semiarid Lands of the West 
North Central States During the First World War," Agricultural History, Vol. 23 (January 
1949), pp. 30-40. 



Nationalizing the American Economy, 1870-1910 / 193 

they were followed by a generation of energetic and ruthless "national business- 
men" who made an indelible impression on the course of American development. 
The Havemeyers in sugar, the Dukes in tobacco, the Weyerhaeusers in lumber, 
the Armours and Swifts in meat-packing, the Carnegies and Fricks in steel, the 
Harrimans and Hills in transportation, the Cookes and Morgans in finance, the 
Rockefellers in oil, became the effective arbiters of national growth. For a 
period, at least, they acquired for themselves and their corporate dynasties huge 
portions of the land, resources, and wealth of the nation. They managed its 
democratic politics in the interest of bounties, franchises, tariffs, and immu- 
nities from taxes or public regulation. If their formal residences were usually on 
the Northeastern seaboard, their creative energies were oriented away from the 
harbors and shipping lanes towards the railroads and freightyards of the in- 
terior. Such men thought and acted in continental terms, for the entire country 
had become a stage on which to play out the continuing drama of their quest 
for wealth. 

To focus only on the achievements of the star performers would be as serious 
a distortion as to neglect the human factor altogether. On every hand were a host 
of lesser lights, smaller figures whose profit horizons were limited to the region 
or locality in which they lived, yet whose strength lay in association or com- 
petition with others of their kind. They were no mere local chorus echoing the 
grand themes of acquisition stated by the principals; they spoke their own lines 
and originated private roles within the continuity of the general plot. They also 
served in opening up the new territories and reaped rich harvests on their per- 
sonal account; their modest capital or credit went into small companies and 
partnerships, into the plant and machinery which heightened the play of tech- 
nology on resources. Then, too, there were the vast majority who made up the 
crowd scenes, both participants and spectators, producers and consumers, who 
took their cues from the main actors or from the contagion around them. As 
they moved west with the surge of population and industry, their expectations 
rose and their efforts were redoubled. Like Tocqueville a half century before, 
Lord Bryce found them in the greater West of the eighties "reaching forward 
to and grasping at the future." 3 

What had occurred since the relatively calm midcentury to release this flood 
of energies in continental development? Did productive possibilities in the West 
of the late nineteenth century differ in any important respect from productive 
possibilities in the West of earlier years? Judged by the type and volume of 
activities, the answer must be that, with few exceptions, it offered little that was 
essentially new; grazing lands, copper ores, petroleum deposits, and great stands 
of timber were by this time hardly novelties. The absolute volume of resource 
inputs, of course, was greatly expanded as a consequence of Western develop- 
All the passionate eagerness of people in the West was directed towards the development 
of their country: "to open the greatest number of mines ... to scatter cattle over a 
thousand hills, to turn the flower-spangled prairies of the Northwest into wheat fields, to 
cover the sunny slopes of the Southwest with vines and olives: this is the end and aim of 
their lives, this is their daily and nightly thought . . ." J. Bryce, The American Common- 
wealth, Vol II (London: Macmillan and Co., 2nd edition revised, 1891), p. 700. 



194 / Regional Economic Development: 1870-1950 

ment. But the greatest gains in output were obtained not from increased quan- 
tities of materials but from the reorganization of production functions along 
capital-intensive lines — that is, by industrialization. Industrialization was a 
process that involved and affected all regions. Multiplication and differentiation 
of areas and activities within the continental system widened and deepened tech- 
nological influences in every aspect of American life; all areas were made acces- 
sible to each other through increasingly refined networks of transportation and 
communications; agriculture was reorganized into specialized "type of farming" 
areas; remote sources of minerals and lumber were made available in order to 
sustain the growth of industrial manufactures. Thus it was the extension of 
industrial technique and organization, not only to manufacture and distribution, 
but to agriculture, mining, and forestry as well, that made Western resources 
economically significant at this point in the nation's history. These resources 
eased the course of industrialization in all parts of the country. But they did 
not determine its pace or its direction; this was a function of demand. 

In early phases of industrialism, the character of demand is usually shaped 
by the requirements of industrialization itself, by the need to accumulate stocks 
of capital goods — plant, machinery, materials, and various forms of "social 
overhead" such as transport services, urban housing, and utilities. In the 
American case, however, many authorities have pointed to the unprecedented 
demands of war after 1860 as imparting a momentum to economic growth that 
sufficed to industrialize the Northeast and carry industrial influences across the 
continent. No doubt, the task of supplying the Union armies with food, clothing, 
and munitions did create opportunities in agriculture, mining, and heavy manu- 
factures; mechanization of some branches of agriculture and some sections of 
the men's garment industry was accelerated. The outcome of the war removed 
certain political obstacles to completion of the transcontinentals and liberalized 
the conditions of public support for important economic interests. Nevertheless, 
even before the war, industrialization had already taken hold in both the New 
England and the Middle Atlantic states and the process of regional differentia- 
tion had gone far. It was probably railroad construction and operation, both 
East and West, that after 1850 gave greatest impetus to economic development 
and set in motion new patterns of resource utilization. 



Technology and Resources: the Role of the Railroads 

The first construction of railroads into the older West had begun as early as 
1853 and, if anything, the prosperity of the later war years attracted capital 
away from what had already proved to be a risky venture. Little new construc- 
tion was attempted during the war years beyond the double-tracking of routes 
into the coal and ore producing districts of the Middle Atlantic states. But the 
manufacture of iron rails alone absorbed almost two-thirds of the increased pig 
iron output during the years 1860-64; and by the end of the war, the total rail- 
way use of iron exceeded 42% of the nation's output. Half the incremental coal 
product of the war years likewise went into fuel supply for rail manufacture or 






Nationalizing the American Economy, 1870-1910 / 195 

was consumed by the locomotives themselves. It was railroad building on this 
scale which quickly revealed critical limits to the industrial raw material poten- 
tial of the Middle Atlantic and New England states. By the end of the war some 
13% of all iron was being derived from oi;es extracted in the Great Lakes 
region; smaller inroads were also being made upon the forests and bituminous 
coal fields of this region in order to meet the demands of industry in the rapidly 
developing Northeast. 4 

The most spectacular surge of railroad construction was reserved for the post- 
war period, when cheaper rails made from Bessemer steel were introduced. In 
1866, rails absorbed less than 1% of the nation's minute steel output and 
steel rails were twice as expensive as iron rails; but in little more than a 
decade, nearly 70% of the greatly enlarged steel output went into rails 
and steel rails cost only one third more than their iron counterpart. Meanwhile 
their superior strength and quality had permitted the development of larger and 
heavier locomotives and had raised the freight potential of the emergent railroad 
network. By the late seventies, also, railroad construction and operation were 
absorbing from 12 to 15% of all coal output and around 10% of the annual 
lumber cut. At the end of the century, when for all practical purposes the great 
continental rail net had been completed, more than 20% of the pig-iron total 
was being rolled into railroad bars. 5 

Railroad construction generated unprecedented demands for capital, land, and 
labor ; larger lines such as the Central Pacific, Union Pacific, and Illinois Central 
employed upwards of 10,000 men each in their peak years of construction. One 
authority estimates that the national figure for railroad construction workers 
probably reached a maximum of 200,000 during the eighties. By that time the 
census reported employment in railroad transportation service in excess of a 
quarter of a million. Similarly, the mobilization of funds for railroad invest- 
ment had become the major concern of financiers on two continents and 
enforced a wholesale adaptation of financial institutions and techniques. Experi- 
ence with railway finance trained a key generation of specialists in the business 
of manipulating huge aggregations of mobile capital. Up to World War I, in 
fact, railroads remained the second largest consumer of capital, second only to 
the construction of buildings for residence and business. 6 Thus, in Leland 
Jenks' words, "the initial impetus of investment in railway construction led in 
widening arcs to increments of economic activity over the entire American 
domain, far exceding in their total volume the original inputs of investment 
capital." 7 

These increments of activity across the nation are the essential contribution 
of the railroad to the dynamics of regional growth. The railroad carried the 

4 P. H. Cootner, Transport Innovation and Economic Development: The Case of the U. S. 
Steam Railroad, unpublished Ph.D. Thesis, M.I.T., (June 1953), Chap. V, pp. 13-14. 

6 Ibid, Chap. VI, pp. 10a and 12. 

6 T. C. Cochran, The American Business System: A Historical Perspective, 1900-1955 
(Cambridge, Mass.: Harvard University Press, 1957), pp. 29, 32-35. 

7 L. H. Jenks, "Railroads As An Economic Force in American Development," Journal of 
Economic History, Vol. 14 (May 1944), p. 7. 



196 / Regional Economic Development: 1870-1950 

industrial revolution into every part of the continental interior. Historians have 
stressed the role of the railroad in reducing the cost of freight and speeding the 
flow of goods, but from the present standpoint at least three other features are 
of equal significance. First, the railroad liberated the course of economic growth 
from the relatively fixed and narrow channels of the coastal and inland water- 
ways systems. The import of this "emancipation" had been evident in some 
areas during the era of sectional rivalry before the war, but in the years after 
1870 its influence was paramount everywhere. Second, the penetration of the 
railroad into the trans-Mississippi West brought a heavy-duty, high-capacity 
transport service into regions of untold resource potential which otherwise could 
have acquired little economic significance. Finally, the physical task of building 
railroads across boundless plains and high mountain passes from the Great 
Lakes to the Gulf and Pacific coasts fostered a growth of heavy manufactures 
and a consequent demand for raw materials beyond the capacity of older 
resource regions to supply. Railroad building in the Plains and Western Great 
Lakes states, for example, necessitated some tapping of the ore, fuel, and forest 
resources of these areas concurrently with their agricultural settlement. 

Railroad mileage in operation during the calendar years 1870, 1880, 1890, 
1900, and 1910 in each of the eight regions is shown in Table 72, together with 



Table 72. Railroad mileage in operation during calendar years and percentage 
increases by decades, by region, 1870-1910 





1870 


1880 


1890 


1900 


1910 






Mileage & 


Mileage & 


Mileage & 


Mileage & 


Region 


Mileage 


% increase 


% increase 


% increase 


% increase 


United States 


52,922 


93,267 


163,596 


193,348 


240,438 






(76) 


(75) 


(18) 


(24) 


New England 


4,494 


5,982 


6,718 


7,521 


7,921 






(33) 


(12) 


(12) 


(5) 


Middle Atlantic 


10,577 


15,147 


19,745 


22.464 


23,777 






(43) 


(30) 


(14) 


(6) 


Great Lakes 


14,701 


25,109 


36,924 


41,007 


44,928 






(71) 


(47) 


(11) 


(10) 


Southeast 


11,843 


16,328 


31,785 


41,134 


55,932 






(38) 


(95) 


(29) 


(36) 


Plains 


8,046 


19,094 


38,354 


42,988 


49,730 






(137) 


(101) 


(12) 


(16) 


Southwest 


711 


4,640 


12,248 


15,302 


25,391 






(553) 


(164) 


(25) 


(66) 


Mountain 


873 


3,236 


9,330 


11,634 


15,550 






(271) 


(188) 


(25) 


(34) 


Far West 


1,677 


3,731 


8,492 


11,298 


17,209 






(122) 


(128) 


(33) 


(52) 



Figures in italics are above the national average. 
Source: Poor's Railroad Manual; Interstate Commerce Commission, Statistics of Railways 
in the U. S. 



Nationalizing the American Economy, 1870-1910 / 197 

percentage increases in track over each intervening decade. After 1880 the bulk 
of the nation's increased facilities appears to have been constructed in the South 
and trans-Mississippi West, although increases in the Great Lakes states remain 
sizable down through 1890. After 1890, however, railroad building falls off 
relatively in the Great Lakes region and also in the Plains. In all Western 
regions the penetration of the railroad seems closely associated with agricultural 
settlement; but in the Mountain and Southwest regions, changes in the rate of 
mileage increase also follow the general increase or decrease in mining employ- 
ment; changes in the general level of business activity seem to have had a greater 
influence on Western railroad building in the nineties than in the seventies. By 
1900, except in the Mountain region, probably few farms were more than ten 
miles or so from a railroad. Meanwhile, the nation was already entering the 
automobile age. 



Differentiation in Agriculture 

The new accessibility of the trans-Mississippi West, together with the growth 
of foreign and domestic markets for farm produce, effected a radical transfor- 
mation in the location of agricultural activities over the decades following the 
Civil War. Already by 1860 the Plains region was contributing about 9% of 
all wheat threshed in the United States, but the heart of the wheatlands still lay 
in Wisconsin and Illinois. A decade later the Plains region's share of wheat 
threshed had risen to 23% and by 1890 reached 37% (see Table 73) . Its share 
of the corn harvested for grain in the United States rose from 21% of the total 
in 1870 to more than 48% two decades later. In 1870 very little milk was sold 
from farms in the Plains, but by 1890 that region (mostly Minnesota and Iowa) 
contributed one-fifth of the greatly expanded national milk supply. Cattle had 
likewise become a major concern in some Plains states by 1870, but by 1890 
the region had more than doubled its share of all cattle population and had over 
27% of the national total. The cattle industry of the Southwest had meanwhile 
received a comparable stimulus as the railroad penetrated further into Texas. 
Over the first half of the decade 1870-80 Texas had also risen to second place 
after Mississippi among the nation's cotton producers, with Arkansas ranked 
third. The spread of the cotton culture into Texas was almost wholly a con- 
sequence of railroad development, since most of the fertile cotton lands were 
far removed from the Gulf coast. 8 

A microcosm of this prolonged agricultural transformation can be found in 
the history of Wisconsin and the adjacent territories. Until the second quarter 
of the nineteenth century, the fertile and partially forested lands between Lake 
Michigan and the Mississippi River had remained isolated and undeveloped. 
Prior to that time their only economic significance had lain in an abundance of 

8 On the relation of transport developments to increased farm production and regional 
marketing problems, see U. S. Industrial Commission, Report on Agriculture and Agricul- 
tural Labor, Vol. 10 (Washington: Government Printing Office, 1901), pp. X-XV, and 
relevant testimony. 



198 / Regional Economic Development: 1870-1950 



Table 73. Regional distribution of selected farm products, 1870-1950 



Wheat (threshed) 



(000's bushels) 



Region 


1870 


1890 


1910 


1930 


1950 


United States 


287,746 


468,374 


683,379 


800,649 


1,006,559 




100% 


100% 


100% 


100% 


100% 


New England 


0.35 


0.07 


0.01 


— 


— 


Middle Atlantic 


14.19 


8.87 


5.98 


4.18 


3.84 


Great Lakes 


44.35 


31.33 


17.74 


12.64 


16.50 


Southeast 


10.30 


8.06 


4.61 


2.44 


2.42 


Plains 


23.40 


37.28 


56.18 


46.83 


38.30 


Southwest 


0.27 


1.01 


2.54 


12.49 


15.86 


Mountain 


0.37 


1.30 


4.16 


12.00 


14.02 


Far West 


6.77 


12.08 


8.78 


9.42 


9.06 




Corn 


(harvested for grain) 
















(000's bushels) 


Region 


1870 


1890 


1910 


1930 


1950 


United States 


760,945 


2,122,328 


2,552,190 


2,130,752* 


2,778,190* 




100% 


100% 


100% 


100% 


100% 


New England 


0.97 


0.29 


0.32 


0.07 


0.03 


Middle Atlantic 


9.81 


3.96 


3.62 


2.94 


2.98 


Great Lakes 


36.49 


27.10 


33.12 


25.09 


35.75 


Southeast 


28.49 


16.91 


16.87 


18.01 


14.63 


Plains 


21.24 


48.23 


39.04 


47.39 


43.73 


Southwest 


2.79 


3.30 


6.71 


5.40 


2.34 


Mountain 


0.04 


0.08 


0.23 


0.98 


0.46 


Far West 


0.17 


0.13 


0.09 


0.12 


0.08 


* excluding silage corn. 




All cattle 








Region 


1870 


1890 


1910 


1930 


1950 


United States 


23,820,608 


57,648,792 


61,803,866 


63,895,826 


76,762,461 




100% 


100% 


100% 


100% 


100% 


New England 


5.71 


2.45 


2.18 


1.98 


1.42 


Middle Atlantic 


16.17 


7.58 


7.41 


6.69 


5.69 


Great Lakes 


22.75 


15.67 


15.89 


16.63 


15.87 


Southeast 


23.93 


15.56 


16.61 


13.95 


17.51 


Plains 


12.32 


27.00 


28.54 


31.12 


28.88 


Southwest 


14.94 


20.24 


17.46 


16.35 


15.99 


Mountain 


0.70 


7.00 


5.99 


7.26 


7.92 


Far West 


3.48 


4.50 


5.92 


6.02 


6.72 



37.51 



28.64 



27.71 



32.08 



27.66 



Nationalizing the American Economy, 1870-1910 / 199 

Whole milk (sold from farms) 

(OOO's lbs.) 



Region 


1870 


1900* 


1910 


1930 


1950 


United States 


2,025,305 


18,359,540 


16,660,400 


38,318,493 


68,529,441 




100% 


100% 


100% 


100% 


100% 


New England 


13.19 


10.04 


9.05 


6.74 


5.00 


Middle Atlantic 


67.07 


32.47 


39.99 


26.94 


20.21 


Great Lakes 


15.63 


28.84 


34.13 


37.54 


37.32 


Southeast 


1.49 


2.18 


2.69 


6.08 


9.01 


Plains 


.86 


20.57 


7.46 


7.35 


11.80 


Southwest 


.03 


0.61 


0.87 


2.80 


3.42 


Mountain 


.07 


1.40 


1.33 


2.61 


2.73 


Far West 


1.66 


3.89 


4.48 


9.94 


10.51 



*1890 figure not comparable. 

Source: U. S. Department of Commerce, U. S. Census of Agriculture: 1950, Vol. II, 
(Washington: Government Printing Office, 1952), pp. 401, 403, 404, 410, 540, 549, 
and 558. 



peltries traded by Indians and in lead deposits recently opened up on some 
scale along Wisconsin's southwestern border with Illinois. Permanent agricul- 
tural settlement had only begun in the years 1834-40, when lands in southern 
Wisconsin were first surveyed and offered for sale. 

Under the existing "frontier" conditions the farmers needed a commercial 
crop within one or two seasons if they were to sustain themselves and their 
families on the new land; and the only staple crop in this part of the country 
was wheat. With great difficulty they sold wheat to the nearby miners and 
woodsmen, to other settlers, to the villages along the lakeshore; much was 
shipped down the Mississippi towards St. Louis and in the early forties exports 
commenced from Milwaukee, then a small town of 2,000 inhabitants. Between 
1849 and 1855 the Wisconsin wheat crop doubled, and by 1860 it had trebled 
again. In 1855 the Milwaukee Board of Trade reported a crop of more than 
8,000,000 bushels and in 1860 a crop exceeding 27,000,000 bushels. The bulk 
of the wheat surplus, moreover, was now shipped to the East. 

The spread of wheat culture across Wisconsin closely paralleled the frontier 
of farm settlement and both were a function of improved communications, by 
river, by plank roads, and finally by the railroad itself. Improvements in farm 
implements during the forties and fifties almost made up for deficiencies in farm- 
ing technique and succeeded for a while in masking the limitation of the soil. 
Meanwhile, the railroads heightened competition from newer wheatlands further 
west, and the combination of price fluctuations and declining yields in the late 
fifties caused serious doubts in older settled sections about the viability of an 
extensive wheat culture. 

Civil War prosperity brought a resurgence of wheat-growing throughout the 
settled parts of the state, though a few of the more enterprising cultivators took 



200 / Regional Economic Development: 1870-1950 

IS70 !9f0 



; % 

50 



1950 



Wheat _.. 



U.S. M.A.G.t S.E, PI S.W. M*. F.W. N.£ M.A. GvL S.S. tt. S.W Mt, F.W. N,E. M.A. <U, IE. Pt. S.W. 



Ozi 



Corn 



N.E. M.A. G.t. SJ, Pf. S.W. M*. F.W. N.E. M.A. GA. S.E W. S.W. Afe, F.W. N.E. M.A. G.L S.E. PI. S.W. Ml. F.W. 



Copper (Mine Production) 



N.E. MA. G.L $.1. Pf. 5.W. Mf. F.W. N.E. M.A, G.i. 5.6. M. S.W. M». F.W. N.E. M.A. G.L S.5. Pi. S.W. M». F.1 



Coal (Mine Production) 




N.E. M.A. G.l. S.E. PI. S.W. M*. F.W. N.E. M-A. O.U S.E. PI. S.W. M*. F.W. N.E. M.A. G.l. S.E. «. S.W. Mt. F.W. 



Pig Ironj 



N.S. M.A. G.l. S.E. Pi- S.W. M». F.W. N.E. M.A. G.L S.E, PL S.W. M*< F.W. N.E. M.A. 6.151. PI, S.W. Ml. F.W. 



F%wre 32. Regional Shares of Production, 1370, 1910, 1950. 



advantage of good times to experiment with other crops, including hops, flax, 
tobacco, and even sorghum. But once the war boom had passed most farmers 
discovered that diversification did not pay. A growing appreciation of the 
state's natural advantages led many to try livestock farming. The first adjust- 
ment was toward sheep-raising which offered most of the remedial, "land-restor- 
ing" features of a grass-grain husbandry at only a fraction of the cost involved 
in a shift into beef or dairying. Further south, in Illinois, resources were shifted 
into the more profitable corn-hog and beef complexes. 

There followed a series of institutional changes, however, under the auspices 
of newly founded regional and state dairymen's associations, which speeded the 
transition to a more complex pattern of animal husbandry after 1870. By 
adopting the "associated dairy system," growing numbers of small farmers in 



Nationalizing the American Economy, 1870-1910 / 201 

northern Illinois and southern Wisconsin found it possible to transfer their 
limited resources into milk production and thereby secure a higher return ; they 
pooled the milk from herds of improved "native" and Shorthorn cattle to pro- 
duce larger amounts of cheese and butter for the local and regional provision 
trade. 

After 1872 remarkable business leadership was provided by the Wisconsin 
State Dairymen's Association, which urged a rejection of "Granger" politics 
and proposed intensive specialization as the more certain road to economic 
survival. Factory cheesemaking was developed, after the earlier pattern of 
New York, as a means of exploiting the special skills of a few expert cheese- 
makers; but equally important vital connections were pioneered in profitable 
dairy markets at home and abroad. While wheat-growing still flourished on the 
newer lands of the western Wisconsin counties and in the newly settled parts of 
Minnesota, the Dairymen's Association sponsored dairy boards of trade and 
promoted reform of farm and factory management in the eastern and southern 
portions of the state in order to enhance Wisconsin's competitive position in the 
great produce markets of Chicago and New York. 

The introduction of fast-freight service and refrigerated cars at this juncture 
facilitated the movement of dairy products over longer distances; a series of 
triumphs at national dairy shows soon secured the state a reputation for prime 
produce. Not all branches of the emerging dairy industry, of course, were 
equally specialized at this time. Farmers who did not yet have cheese factories 
in their vicinity were advised to make butter and since this was a relatively 
simple process there was less need for skilled operatives or factory organization. 
As a consequence, Wisconsin buttermaking remained a domestic chore until 
well into the eighties when technical advances in cream separation combined 
with price changes in regional dairy markets to foster a rapid growth of cream- 
eries. Since Wisconsin had large pre-empted leadership in cheesemaking, 
dairymen in the adjacent states of Minnesota and Iowa concentrated their first 
efforts on buttermaking. Meanwhile in Illinois the expansion of the Chicago 
milk shed caused many farmers to direct their output into the more profitable 
market for fluid milk. 

Thus the revolution in dairy organization, technique, and markets gradually 
disrupted the older modes of production and enlarged the element of skill in 
the new. Under a double compulsion to preserve the fertility of the soil and 
raise the return on its use, dairy farmers of Wisconsin and adjacent areas con- 
centrated on the improvement of feeds and breeds in order to achieve an in- 
creased flow of milk at lower costs per unit. This was accomplished between 
1870 and 1890 by a more balanced crop program and by extending lactation 
through the winter months. Ensilage (silos) provided the long-sought solution 
to winter dairying problems while research, at the State University and in other 
parts of the country, related the cow's yield of milk to her intake of digestible 
nutrients. Further experiment, based on German examples, resulted in the 
development of a "balanced dairy ration" which combined the several feed 
components in the correct proportions required by the flow of milk. Finally in 
1890 the celebrated Babcock fat-test provided at once the first cheap, accurate 



202 / Regional Economic Development: 1870—1950 

measure of milk quality and a more equitable basis for making payments for 
milk at the factory; it compelled even closer attention to scientific breeding and 
better feeding practices. 9 

The continuing revolution in dairy farm management led, in turn, to more 
numerous and larger manufacturing establishments. A series of technical and 
scientific developments permitted even greater control over manufacturing 
processes. By the nineties, improvements in milk sanitation and pasteurization 
helped reduce the hazards of fluid milk consumption, condenseries had obtained 
a foothold in northern Illinois and southeastern Wisconsin, and hence a growing 
share of the regional milk supply was being diverted to lucrative market milk 
outlets in midwestern cities. Milk sold as whole milk or cream to city milk 
dealers (or condenseries) generally displaced milk manufacturing (cheese and 
butter) in areas close to large urban populations, where land values and most 
other cost factors (except transportation) were usually high. Cheese and butter 
manufactures slowly relocated in belts beyond the market milk zones. In Wis- 
consin, climatic conditions and resultant crop programs usually gave milk sold 
for cheese an edge over milk sold for butter, but in southwestern parts of the 
state and in adjacent sections of Iowa and Minnesota, butter-making in con- 
junction with hog-raising and corn-growing proved more viable. 10 Thus com- 
petition for milk among the different branches of dairying gradually enforced 
a pattern of local specialization within the dairy belt of the Great Lakes and 
Plains regions as it had done somewhat earlier in the more populous Middle 
Atlantic and New England regions. 11 

Table 74 highlights the continuing relocation of agricultural production in 
the late nineteenth and early twentieth centuries. The wheat belt gradually 
moved further west, eventually settling in the western Plains, Southwest, and 
Mountain states and becoming increasingly dissociated from the corn belt. 
Meanwhile livestock activities concentrated more intensively in the rich corn 
belts and grasslands of the Great Lakes and eastern Plains states. The corn- 
hog-cattle complex, for example, located in southern parts of these regions where 
growing seasons are notably longer, while the grass-cattle-dairy complex tended 
to concentrate north of the corn-belt. Local adjustments within this over-all 
pattern have occurred from time to time during the present century but they do 
not appear related to the expanding transportation system or, for that matter, 

9 Henceforth cows were tested for their individual yields and herds were slowly up-graded. 
By 1900, Holstein and Channel Islands' breeds were preferred above all others; public and 
private efforts were joined to eliminate the scourge of bovine tuberculosis from Wisconsin 
herds; the interrelation of technical and economic forces in the growth of dairying is dis- 
cussed at length in the forthcoming volume, E. E. Lampard, The Rise of the Dairy Industry: 
A Study of Agricultural Change in the Midwest, 1820-1920, (Wisconsin State Historical 
Society, Madison, Wis.). 

10 Skim milk from creameries, or separated on the farm, provided an important supplement 
to hog feed. Corn quickly became the staple animal feed, hence beef cattle were also well- 
suited to states in which corn was the principal grain. Corn-hog-beef production at this time 
was usually more profitable than dairy farming, but its geographical spread was limited by 
the exacting climatic requirements of corn culture. 

"E. Brunger, "Dairying and Urban Development in New York State, 1850-1900," Agricul- 
tural History, Vol. 29 (October 1955), pp. 169-74. 



Nationalizing the American Economy, 1870-1910 / 203 



Table 74. Leading states in the production of whole milk sold from farms, 
wheat threshed, and all cattle, 1870-1950, as a percentage of U. S. total 

IT hole milk 



187C 


1 


1900 


* 


1910 


1930 


1950 


N. Y. 


57.65 


N. Y. 


20.85 


N. Y. 


27.06 


Wis. 


18.93 


Wis. 


18.47 


Ohio 


9.46 


Wis. 


11.82 


Wis. 


15.34 


N. Y. 


15.32 


N. Y. 


10.92 


Mass. 


6.49 


Iowa 


10.03 


Pa. 


8.75 


Pa. 


8.23 


Cal. 


7.64 


Pa. 


6.12 


111. 


8.74 


111. 


8.16 


Cal. 


6.67 


Pa. 


6.30 


111. 


3.93 


Pa. 


8.01 


Ohio 


5.13 


111. 


5.35 


Minn. 


5.61 


Conn. 


2.66 


Minn. 


4.86 


Mich. 


3.82 


Ohio 


5.09 


Ohio 


5.26 


Wheat threshed 


1870 


> 


189C 


) 


1910 


193C 


> 


1950 




111. 


10.47 


Minn. 


11.17 


N. D. 


17.08 


Kans. 


18.55 


Kans. 


14.47 


Iowa 


10.23 


Cal. 


8.73 


Kans. 


11.34 


N. D. 


11.94 


N. D. 


10.49 


Ohio 


9.69 


111. 


7.98 


Minn. 


8.34 


Neb. 


6.73 


Okla. 


7.82 


Ind. 


9.64 


Ind. 


7.97 


Neb. 


6.98 


Okla. 


6.39 


Tex. 


7.48 


Wis. 


8.90 


Ohio 


7.59 


S. D. 


6.89 


Tex. 


5.51 


Wash. 


5.73 


Pa. 


6.84 


Kans. 


6.49 


Wash. 


5.99 


Wash. 


5.32 


Ohio 


5.27 


All cattle 


1870 


i 


1890 


i 


191C 


> 


1930 




1950 


i 


Tex. 


14.68 


Tex. 


14.83 


Tex. 


11.22 


Tex. 


10.33 


Tex. 


10.20 


N. Y. 


8.59 


Iowa 


8.49 


Iowa 


7.20 


Iowa 


6.48 


Iowa 


5.92 


m. 


7.20 


Kans. 


5.50 


Kans. 


4.98 


Wis. 


5.54 


Wis. 


4.88 


Ohio 


6.03 


111. 


5.31 


Neb. 


4.74 


Kans. 


5.05 


Neb. 


4.73 


Pa. 


5.64 


Mo. 


5.15 


Wis. 


4.34 


Minn. 


4.94 


Kans. 


4.57 


Mo. 


4.84 


Neb. 


3.72 


Mo. 


4.14 


Neb. 


4.93 


Mo. 


4.27 






♦Figures for 1890 not available. 
Source: U. S. Census of Agriculture, 1950, Vol. II, pp. 401, 410, 549, 558. 



to the taking-up of new land. They have more to do with changes in demand, 
wartime needs for food and fibers, and government policies for agriculture. 
Nevertheless, some adjustments have stemmed from improvements in technique 
— the introduction of new breeds of cattle (important in the South), the devel- 
opment of hybrid corn (important in areas with short growing seasons), and 
the extension of irrigation (important in semi-arid areas of the Southwest and 
Far West). 12 

12 L. Haystead and G. C. Fite, The Agricultural Regions of the United States (Norman: 
University of Oklahoma Press, 1955) ; E. S. Dunn, Jr., The Location of Agricultural Pro- 
duction (Gainesville: University of Florida Press, 1954). 



204 / Regional Economic Development: 1870-1950 



Differentiation in Mining Activities 

Regional developments in mining activities in the late nineteenth century 
were, perhaps, even more spectacular than those in agriculture. At the close of 
the Civil War, for example, the Great Lakes region was providing about 13% 
of the nation's iron and a somewhat larger proportion of its iron ore. By 1870 
some 30% of the iron ore total originated in this region, and by 1890 the pro- 
portion had risen to 48%. 

Over the same period the mining of Alabama-Tennessee ores raised the South- 
east's share of national ore production from 2% to almost 20%. It was no coin- 
cidence that this great expansion in Southern mining, like that in the Great 
Lakes states, accompanied two decades of rapid railroad building both region- 
ally and nationally. In the Southeast, moreover, it was a genuine regional 
expansion in the sense that, although much of the enterprise and capital came 
from outside, much of the demand for iron products originated in the region 
itself. For a period, at least, the Southern iron manufacture benefited from its 
low-cost production, which ranged as much as 26% below Pittsburgh. The ores 
were not suited to the Bessemer process of steelmaking, yet their proximity to 
cheap coal and improved access to growing markets for foundry iron sufficed to 
insure their lively development. 

In contrast to the Great Lakes and Southern iron fields, the great ore reserves 
of the northern Plains were largely unaffected by the national demand for iron 
and steel products. As late as 1890 the entire Plains region supplied less than 
8% of the nation's ore requirements. There were, to be sure, certain processing 
problems connected with the soft and crumbly hematite of Minnesota which 
probably retarded Plains development; on the other hand, the ore was admirably 
suited to the newer stripping and steam shovel techniques used in selective 
mining. It required only the linking of the local Mesabi mine railways to the 
Great Northern system in the late nineties to release this ferric treasure for 
American industry. 13 Unlike the earlier exploitations of iron ore in the Middle 
Atlantic, Great Lakes, and Southeast regions, however, the rapid development of 
the Minnesota ores did not result in any great upsurge of local pig-iron produc- 
tion. The proximity of the mines to cheap water routes on the Great Lakes made 
these new fields a tributary to the iron and steel manufacture of the Great 
Lakes and Middle Atlantic states. 14 

Lead mining was perhaps even more profoundly affected by railroad develop- 
ment than iron. In 1870 about 60% of the lead produced in the United States 

13 J. W. Thompson, "The Genesis of the Great Northern's Mesabi Ore Traffic" Journal of 
Economic History, Vol. 16, (December 1956), pp. 551-57. In 1906 James J. Hill claimed 
that "never in the whole world has the same amount of track moved the same tonnage, or 
approached it." Between 1896 and 1900 shipments of iron ore from Lake Superior ports 
doubled, from 9,600,000 tons to 18,600,000 tons: 80 per cent of this ore was received by ten 
ports on the lower shores of Lake Erie; Industrial Commission, Vol. 19, op. cit., pp. 472-74. 

"The opening of the Lake Superior ore district appears to have coincided with the 
abandonment of the high-cost production of the Appalachian range. Experts of the U. S. 
Geological Survey estimated that, at the prevailing rate of output of the early 1900's, produc- 



Nationalizing the American Economy, 1870-1910 / 205 

came from the Great Lakes region and about 34% from newer developments in 
the Plains. At this time no significant amounts of lead were reported from the 
Mountain region. But a decade or so later, after branch lines from the mines 
had linked up with major systems, Colorado, Utah, and Nevada supplied more 



Table 75. Domestic mine production of recoverable lead,* by region, 1870- 
1950 

(1870 = current dollar value; short tons thereafter) 



Region 


1870 


1890f 


1910 


1930 


1950 


United States 


$736,004 


181,141 


386,899 


556,948 


430,678 




100% 


100% 


100% 


100% 


100% 


New England 


.41 


— 


— 


— 


— 


Middle Atlantic 


1.02 


— 


— 


.50 


.34 


Great Lakes 


60.10 


1.02 


1.21 


.32 


.76 


Southeast 


3.94 


.16 


.05 


1.03 


.80 


Plains 


34.26 


26.62 


41.35 


38.16 


33.46 


Southwest 


— 


4.37 


1.63 


6.80 


11.93 


Mountain 


— 


66.70 


54.61 


50.70 


44.44 


Far West 


.27 


1.13 


1.15 


2.49 


8.27 



*Less zinc residues. 

tl81,141 short tons is a combined total of production of lead-bearing ores in Rocky 
Mountain states and the lead recovered in other states. The total recovered lead reported 
in the Mineral Industries section of 1890 U. S. Census, pp. 163-173, is 182,967 short tons. 

Source: 9th U. S. Census, 1870, Vol. Ill, p. 768; Compendium of the 11th U. S. Census: 
1890, Pt. 2, pp. 479-80; Mineral Resources, 1910, Pt. I, p. 234; Mineral Resources, 
1930, p. 489; Minerals Yearbook, 1951, p. 742. 



than 66% of the enlarged national lead requirements. 15 Meanwhile production 
of recoverable lead in the Great Lakes region had declined to a trickle. In the 
Plains region, on the other hand, Missouri has continued as a major producer of 
lead down to the present; and since 1910 Far Western production has assumed a 
modest importance as older regions have declined. 

That the "new accessibility" was not the whole story of Western mineral 
development in the late nineteenth century, however, is well illustrated by the 
history of copper mining. Penetration of new territory by the railroad did not 
at once guarantee the exploitation of its mineral wealth. Thus the fact that 
"lake copper" from Michigan adequately met requirements before the eighties 
delayed the construction of feeder lines into the Montana and Arizona copper 

tion would reach a maximum in the next thirty years and begin to decline; National 
Conservation Commission, Report, (3 vols., Washington: Government Printing Office, 1909), 
Vol. I, pp. 95-110; Vol. Ill, pp. 483-520. If anything, their concern for the coal supply at 
the time was even greater, since, unlike the iron, fuel could only be used once; ibid, Vol. 
Ill, pp. 426-446. 

15 U. S. Industrial Commission, Vol. 19, op. cit., pp. 229, 242-46; W. R. Ingalls, Lead and 
Zinc in the United States, (New York: Hill Publishing Co., 1908). 



206 / Regional Economic Development: 1870-1950 

lands, though their potential was publicized by seekers after the precious metals 
in the 1860's. It was an important shift in the utilization of copper following the 
rise of the electrical industries which revealed the essentially inelastic conditions 
of supply in older mining areas and compelled recourse to the new. 16 During the 
eighties the market price of copper fell off somewhat, but smelter production 
increased by more than 300% ; both changes can be attributed to the impact of 
ores from the Mountain and Southwest regions. 17 

But if improved transportation and shifts in demand combined to make 
Western ores available to meet national requirements, many technical problems 
had to be solved before the mineral resources of newer regions could be fully 
developed. The mining of ores under the wide range of geological conditions 
prevalent in the greater West taxed the accumulated experience and ingenuity of 
mining engineers and metal processors in two continents. New and deeper 
understanding of ore genesis and location stemmed directly from the work of the 
United States Geological Survey (1879) and later from such private ventures 
as the Anaconda Geological Department (1898). In turn, this intelligence led to 
more careful planning and preliminary development, while new and complicated 
machines gave added power to the grossly inadequate supply of skilled labor. 
The introduction of improved tools and better explosives and the perfection of 
mechanized pumping, drilling, hoisting, and handling equipment required the 
massing of great funds of mobile capital for Western mining enterprise in much 
the same fashion as for the railroad. 18 

Copper, once more, provides impressive evidence of the creative interplay of 
technology on resources. The rich "native" ores from the Great Lakes region 
were notably free from impurities, and comparatively simple concentration 
processes sufficed to yield high-grade metal suitable for all existing uses of cop- 
per and its principal alloys. Likewise, the first ores extracted from the Butte 
mines in Montana were so rich that they required no exacting concentration. 

16 Copper had long been used for sheathing ships but steel was substituted for this purpose 
during the nineties; it had also been used for roofing and gutters as well as for domestic 
and laboratory utensils (brass). "The great development of electricity in recent years . . . 
has added very largely to the demand for copper, chiefly on account of its conductivity, 
and it is not likely that any cheaper material can be substituted with satisfaction." U. S. 
Industrial Commission, Vol. 19, op. cit., pp. 230-32. This source also has an interesting 
commentary on the differences in mining laws between Michigan and Montana. 

"The same decade significantly saw the beginnings of local coal mining in the Mountain 
region, chiefly in Colorado and Wyoming. Coal mines were also developed in Utah. In 
competition with Wyoming mines, Colorado and Utah produced for the railroads and for 
local smelters. Some producers also attempted to sell anthracite in California but ran into 
stiff competition from sea-borne coal carried as ship's ballast; U. S. Industrial Commission, 
Report on Capital and Labor Employed in the Mining Industry, Vol. 12 (Washington: 
Government Printing Office, 1901), p. VII. 

18 A summary of technical advances in the mining industries and metal extraction during 
this period is given in U. S. Industrial Commission, Vol. 19, op. cit., pp. 235-50; also 
C. Kirchoff, "A Decade of Progress in Reducing Costs," in American Institute of Mining 
Engineers, Transactions, Vol. 29 (1899), pp. 352-71; and C. C. Spence, "When the Pound 
Sterling Went West: British Investments and the American Mineral Frontier," Journal of 
Economic History, Vol. 16 (December 1956), pp. 482-92. 



Nationalizing the American Economy, 1870-1910 / 207 

But within a few years recourse to deeper sulphide ores, averaging less than 6% 
copper, rendered older crushing and gravity concentration methods increasingly 
impractical. 19 Some electrolytic refining was also introduced during the nineties 
and produced an excellent Western copper at relatively low cost. But the great 
changes became necessary after 1905 with the shift from the mainly oxidized 
ores to so-called "porphyry" sulphide deposits assaying between 2.5% and 0.75% 
copper. Their development was facilitated in part by the adaptation of selective 
mining techniques from the Mesabi iron range but in much greater part by the 
"oil flotation" method of concentration which gave greater recovery from the 
same grade of ore than the old fashioned gravity method. Selective flotation 
made it possible to concentrate practically all Western ores and even tailings, 
although the variety of local conditions still imposed special process features on 
particular operations. 20 Already before World War I, the combined output of six 
large porphyry mines in the West made up nearly 30% of the nation's copper 
supply. 

The significance of these technological advances which (with the possible 
exception of the leaching of oxidized ores) were all in general use by the second 
decade of this century, can be seen in a comparison of the declining yields from 
various United States ores. During the heyday of the Calumet and Hecla Mining 
Company in Michigan before 1880, native ore had once run as high as 20% 
copper. But in the years 1887-1905, the average yield of "lake ore" was only 
2.96%, somewhat below the country-wide average of about 3% during the 
nineties. By 1906 the national average yield had declined to 2.5% and by 1915 
to 1.66%. A decade later the figure was 1.54%, and by the midcentury it had 
dropped to 0.91 percent. 21 

The advent of deeper mining and the growing capital requirements for both 

19 Steam stamping proved very wasteful and was quickly replaced by a staged crushing 
and rolling treatment with concentration at each successive stage in order to remove the 
bulk of copper and other concentrates as they broke away from the gangue and were 
reduced to slimes. The finer copper matte was then smelted in blast and, to obtain still 
higher yields, in reverberatory furnaces or modified Bessemer Converters (blister copper). 

20 By oil flotation methods, the ore, after some preliminary milling, is treated in a mixture 
of water and chemicals (chiefly oils) through which air is driven to carry the heavier 
minerals to the surface in bubbles. The resulting froth is then removed and the oils sep- 
arated from the metal concentrate by heat. 

21 U. S. Department of the Interior, Bureau of Mines, Materials Survey — Copper, (1952), 
p. VI-41. Mine sources of U. S. copper production in 1950 are shown in the following 
table, ibid, p. 11-25 : 

Ore Per cent Per cent of 

(short tons) yield total copper 



Concentrating ores: Native 


(Mich.) 


4,386,474 


0.58 


2.9 


Sulphides and oxides 




85,819,695 


0.89 


86.5 


Smelting ores 




624,261 


3.37 


2.4 


Leaching ores 




3,755,362 


0.88 


3.7 


Copper from precipitates 




— 


— 


4.5 



100.00 



208 / Regional Economic Development: 1870—1950 

the extracting and smelting phases of copper production tended to increase the 
size of the major producers. For example, through large investments of New 
England capital, Calumet and Hecla had acquired a dominant position in Michi- 
gan mining before the Civil War, and in the highly competitive period following 
the introduction of Western ores, economies of large-scale operations proved 
decisive in both the old areas and the new (Table 76) . 22 The same tendency was 



Table 76. Domestic copper mine production, by region 1870-1950 

(1870 — current dollars; short tons thereafter) 



Region 


1870 


1890 


1910 


1930 


1950 


United States 


15,201,312 


111,715 


541,998 


688,748 


909,337 




100% 


100% 


100% 


100% 


100% 


New England 


6.90 


.03 


— 


— ■ 


— 


Middle Atlantic 


1.52 


— 


.08 


.21 


— 


Great Lakes 


82.90 


39.46 


20.54 


12.29 


2.82 


Southeast 


7.96 


.01 


1.57 


— 


— 


Plains 


— 


— 


— 


.01 


.33 


Southwest 


.14 


15.79 


27.87 


46.57 


51.64 


Mountain 


— 


44.63 


39.49 


28.34 


37.21 


Far West 


.58 


.08 


10.44 


10.01 


6.41 


Unallocated 


— 


— 


.01 


2.57* 


1.59f 



* Figure given is for North Carolina, Tennessee, and Vermont. 
tCovers Pennsylvania, Tennessee, and Vermont. 

Source: 9th U. S. Census, 1870, Vol. Ill, p. 767; Mineral Resources 1910; Pt. I, pp. 172, 170 
(Mine Returns); Mineral Resources 1930, Pt. I, p. 706; Minerals Yearbook 1950, 
p. 473. 



soon apparent in the West as one or another large corporation came to dominate 
its locality; eventually both mining and smelting processes were integrated 
under single managements. Of the four or five giant mining companies of the 
present century, however, only Anaconda moved very far into the fabricating 
side of the industry through its subsidiary, American Brass, leaving American 
Smelting and Refining, not a mining company, to take the lead in refining 
phases of the industry. 

Not only did the development of mining corporations and mineral technology 
facilitate the opening up of new fields in the West, but the special geological 
features of each area prompted innovation in technique and additions to scien- 
tific knowledge. L. D. Graton has summarized this interaction between experi- 

22 W. B. Gates, Michigan Copper and Boston Dollars: An Economic History of the Michi- 
gan Copper Industry (Cambridge: Harvard University Press, 1951); R. G. Raymer, 
A History of Copper Mining in Montana (Chicago: The Lewis Publishing Co., 1930) ; 
U. S. Industrial Commission, op. cit., (1901), p. 606; ibid., Vol. 19, op. cit., pp. 232-35, for 
the formation and operations of the Amalgamated Copper Company. 



Nationalizing the American Economy, 1870-1910 / 209 

ence and understanding through which, one by one, the mining districts "yielded 
up specific, fundamental secrets" as follows: Leadville (1882), replacement and 
certain aspects of structural control; Marquette (1893), structural barriers for 
descending solutions; Butte (1896), secondary sulphide enrichment; Seven 
Devils (1899), contact-metamorphic ores; Rico (1901), structural barriers for 
ascending solutions; Coeur d'Alene (1908), pre-ore faulting of vain structures; 
Goldfield (1909), alunitic alteration 23 



Differentiation in Iron and Steel Production: the Dominance of Fuels 

The dynamic influences of science and technology were not, of course, con- 
fined to the processes of mining and smelting ore; they pervaded almost every 
branch of productive activity. The record of iron and steel, thought by many 
observers of the late nineteenth century to be the epitome of economic progress, 
reveals the influence of technology on processing and fabricating to be at least 
as significant as on the extraction and treatment of ores. If the development 
of the Minnesota iron ranges marked a profound shift in the regional pattern of 
ore supplies, changes in the location of iron and steel production were affected 
by other forces, more especially by developments in metallurgical science and 
fuel supply. Compared with either lead or copper, iron production was not 
critically dependent on the provision of new railroads. 24 Thus, as Table 77 
shows, the regional pattern of iron manufacture, apart from expansion in the 
South, was not drastically changed down to World War I (or, for that matter, 
thereafter). Regardless of shifts in ore production, the Middle Atlantic and 
Great Lakes regions dominated pig-iron production throughout the period 1870- 
1910 and have declined only moderately since, notwithstanding the recovery 
of the South and more recent developments in the Mountain and Far Western 
regions. The impact of Mesabi ores after 1890 was felt not in the Plains or 
even the western Great Lake states, but in Ohio and other manufacturing centers 
linked to the ports along the southern shores of Lake Erie. 

In steelmaking the growing ascendancy of the Great Lakes region within the 
larger Northeastern industrial belt was already apparent before 1910, though 
its rise to dominance was delayed until the interwar years. The attraction of the 
cheap water route from Minnesota was no doubt a factor in the decisions of steel 

23 L. D. Graton, "Seventy-five Years of Progress in Mining Geology," Seventy-five Years of 
Progress in the Mineral Industry 1871-1946, (New York: The American Institute of Mining 
and Metallurgical Engineers, 1947, p. 10. Also, C. R. Hall, History of American Industrial 
Science (New York: Library Publishers, 1954), pp. 185-216. 

2t Iron ore moved over some 200 miles of railroad, in addition to its voyage on the Lakes, 
to reach the furnaces of the manufacturing belt. Coal, limestone, ore, and pig loomed large 
in the bulk of railroad freight in the Great Lakes region. Until 1860 Pennsylvania was the 
largest iron ore producer, but by 1902 it had fallen to sixth place among the states. Iron 
manufactures in the Atlantic coastal states benefited from their seaboard location, and ore 
imports rose rapidly in the late nineteenth century. By 1913 they exceeded 2,500,000 tons, 
most of which came from Cuba and Chile. Isaac Lippincott, Economic Development of the 
United States (New York: D. Appleton and Co., 1921), pp. 176, 330-32. 



210 / Regional Economic Development: 1870-1950 
Table 77. Pig iron production, by region, 1870-1950. 



(long tons) 



Region 


1870 


1890 


1910 


1930 


1950 


United States 


l,832,878 a 


9,202,703 a 


27,303,567 a 


31,752,169 a 


57,666,937 




100% 


100% 


100% 


100% 


100% 


New England 


1.68 


.28 


.06 


— 


— 


Middle Atlantic" 


66.53 


54.89 


50.55 


39.42 


34.90 


Great Lakes c 


25.44 


27.13 


36.57 


44.36 


43.94 


Southeast* 


6.35 


17.35 


11.25 


13.69 


17.05 


Mountain 6 


— 


.23 


— 


2.53 


— 


Far West* 


— 


.12 


1.57 


— 


4.11 



a Includes ferro-alloys. 
b Includes Massachusetts after 1910. 
c Includes Missouri in 1890, Minnesota in 1910 and 1950. 
d Includes Texas after 1870 and Maryland in 1950. 
e Includes Wisconsin, Minnesota and Iowa in 1930. 

1 Includes small amounts in Missouri and Colorado in 1910, and Colorado and Utah in 
1950. 

Source: U. S. Census 1870, Vol. Ill, p. 603; American Iron and Steel Association, Annual 
Statistical Reports, 1890, p. 62 and 1910, p. 95; American Iron and Steel Institute, 
Annual Statistical Reports, 1930, p. 4 and 1950, p. 18. 



men to move the Lackawanna plant from Scranton to Buffalo (1901) and to 
locate the new Gary works in Indiana, as it was in the still earlier proliferation 
of Bessemer plants around Chicago. However, more potent factors can be found 
in the superior access of Lake sites to the westward-moving steel market (Buffalo 
also provided cheaper access to some Eastern markets) and in the growing com- 
plications of the fuel supply. The persistence of an iron and steel complex in 
southern Illinois and eastern Missouri after the virtual impoverishment of local 
ores is further testimony to the importance of fuel and market considerations in 
the steel geography of the period before 1910. 

Over the decades 1870-1890 larger and more efficient blast furnaces raised 
average furnace output sixfold, while the Bessemer converters multiplied in 
areas with access to non-phosphoric ores. But already in the seventies the indus- 
try was confronted with the critical inadequacy of its principal charcoal and 
anthracite fuels. Uncoked bituminous coals had been used in Pennsylvania 
during the 1840's and many new furnaces in the eastern Great Lakes states had 
also adopted this fuel, but conversion to coke (often in conjunction with anthra- 
cite) was not common until the Civil War decade. By 1871 about 30% of the 
total iron output was produced with so-called "beehive" coke, and within the 
next few years the proportion produced with coke fuel surpassed the combined 
proportion produced with anthracite and charcoal. 25 

^Charcoal had remained the principal American smelting fuel until about 1855 and the 
persistence of "obsolescent" technique is underlined by the fact that as late as 1917 some 



Nationalizing the American Economy, 1870-1910 / 211 

After 1859 Connellsville coke and Lake ores provided the material bases for 
the rise of western Pennsylvania, and notably Allegheny county, to leadership in 
the American iron industry. From the mid-90's, coke from "by product" ovens 
began to displace the beehive coke; nevertheless, it was not before World War I 
that the steel industry converted to by-product fuel on any large scale. 26 By 
1919 by-product ovens provided 57% of all coke, and the output of the beehive 
variety went into steady decline. In view, therefore, of the appreciably greater 
weight of fuel than ore required in the manufacture of steel ingots before 1910, 
it is not surprising that major blast furnaces were located near sources of coking 
coal. 27 Without improvements in fuel technology, which between 1879 and 1954 
reduced the fuel consumption per ton of pig by 40%, it is unlikely that iron 
and steelmaking would have been released from its century-long tie to coal. 28 

If the volume of iron and steel output was ultimately dependent on the availa- 
bility of ore, fuels (Table 78), and alloy ingredients, the actual yield might vary 
with the process of manufacture. Around 1870, for example, only about 80,000 
net tons of steel ingots were produced in the entire country, of which almost 
45% was obtained by antiquated crucible, puddled iron, and other processes. 29 
Not before the eighties did the improved Bessemer process begin to make its 

376,525 tons of pig were made in 18 blast furnaces using charcoal fuel. American Iron & 
Steel Institute, Annual Statistical Report, 1917, p. 21. 

26 The utilization of coal-chemical by-products (ammonia, coal tar, and benzol) made an 
important addition to revenues of the iron and steel industry, especially after the outbreak 
of war in 1914 had closed off German sources of supply. 

"Approximate consumption of coke per ton of pig-iron produced in the United States, 
1879-1954: 

(tons) 
Coke* Coal Equivalent 

1879 1.47 2.10 

1889 1.29 1.85 

1899 1.20 1.72 

1909 1.13 1.62 

1919 1.07 1.53 

1929 .91 1.31 

1939 .88 1.27 

1954 .87 1.24 

* Coal-Coke conversion factor = 70%. 

Source: Walter Isard, "Some Locational Factors in the Iron and Steel Industry since the 
early 19th Century," Journal of Political Economy, Vol. 56 (1948), p. 205; American 
Iron and Steel Institute, Annual Statistical Report, 1940, p. 11, 1956, p. 46. 

2S Control of apparently minor difference in fuel input, air infiltration, and draft regulation, 
mostly introduced since 1910, have been the principal factors in reducing fuel consumption 
from upwards of 6 million B.T.U. per ton of ingots to between 3-4 million B.T.U. character- 
istic of the best recent practice: C. D. King, "Seventy-five Years of Progress in Iron and 
Steel," Seventy-Five Years of Progress in the Mineral Industry, 1871-1946 (New York: The 
American Institute of Mining & Metallurgical Engineers, 1947), p. 188. 

29 Some 45,000 net tons of ingots were made in six Bessemer plants located in the Middle 
Atlantic and Great Lakes regions (3 in Pennsylvania, and 1 each in Michigan, Ohio, and 
Illinois) ; 2,000 net tons were made by the one open-hearth plant located in New Jersey. 
(Ibid., pp. 174-82.) 



212 / Regional Economic Development: 1870-1950 

weight felt in American steel production after the incorporation of the improved 
Thomas-Gilchrist "basic" acid lining (1878). By 1890 the Bessemer processes 
contributed about 80% of the 5,000,000 net tons of steel ingots made in the 
United States. 30 



Table 78(a). Coke production, by region, 1880-1950 

(net short tons of beehive and by-produce oven coke) 



Region 


1880 


1890 


1910 


1930 


1950 


United States 


3,338,300 


11,508,000 


41,709,000 


47,972,000 


72,718,000 




100% 


100% 


100% 


100% 


100% 


New England 


_ 


— 


1.08 


2.85 


2.02 


Middle Atlantic 


84.51 


74.38 


66.06 


42.69 


42.33 


Great Lakes 


3.39 


.96 


5.79 


37.13 


35.09 


Southeast 


11.16 


22.03 


21.50 


13.41 


14.13 


Plains 


.09 


.16 


.74 


2.38 


1.99 


Southwest 


.05 


.08 


1.21 


— 


.94 


Mountain 


.80 


2.34 


3.47 


1.44 


2.79 


Far West 


— 


.05 


.14 


.10 


.71 



Source: Mineral Resources, 1920, Part II, folder inserted in back of book; Mineral Re- 
sources, 1931, Part II, pp. 382-83; Minerals Yearbook, 1950, p. 410. 



Meanwhile the somewhat cheaper, though slower, open-hearth (Siemens- 
Martin) process had been introduced from Europe and, with improved linings, 
had shown itself suited to all grades and types of ore. 31 Its slowness, moreover, 
was soon recognized as an advantage over the Bessemer process since it facili- 
tated much greater qualitative control over the final steel product. A second 
great advantage of the open-hearth is its ability to take scrap metal as part of its 
charge, a factor which told increasingly in its favor during the twentieth century. 
Finally the open-hearth is a much smaller operation than the Bessemer, and the 
surge of small plants towards the close of the century was mostly a consequence 

30 The core of Bessemer technique involved blowing air through the molten metal in the 
converter, thereby oxidizing part of the iron and most of the manganese, silicon, and carbon 
contents. Sulphur and phosphorus, however, were unaffected by the oxygen and adoption of 
the process was largely dependent on the availability of low phosphoric ores. The Thomas- 
Gilchrist basic lining absorbed most of the residual phosphorus in the converter and made 
possible the wider use of phosphoric ores. The first use of an improved magnesium lime- 
stone base in the place of the acid bottom came in 1884 but, according to C. D. King, the 
most important "single development in steel," came after 1889 when "the hot metal mixer 
came into general use (and) the Bessemer became firmly established on a large scale." 
Ibid., p. 181. 

31 The Lakeside plant of the Otis Company in Cleveland is said to be the first establish- 
ment designed for the exclusive manufacture of open-hearth steel in the United States 
(1874) and it was here in 1886 that the first basic open-hearth operation was conducted 
with magnesite imported from Austria. 



Nationalizing the American Economy, 1870-1910 / 213 

Table 78(b). Coal production, by region, 1370-1950 

(short tons of bituminous, lignite, and anthracite coal) 



Region 


1870 


1890 


1910 


1930 


1950 


United States 


33,035,580 


157,770,963 


501,596,378 


536,911,000 


559,976,000 




100% 


100% 


100% 


100% 


100% 


New England 


— 


— 


— 


— 


— 


Middle Atlantic 


76.53 


58.39 


47.89 


36.52 


26.90 


Great Lakes 


17.01 


19.12 


19.95 


17.40 


20.35 


Southeast 


3.02 


11.27 


21.56 


38.34 


46.60 


Plains 


2.78 


5.73 


3.23 


2.22 


1.84 


Southwest 


— 


.91 


1.61 


1.05 


.61 


Mountain 


.18 


3.67 


4.97 


4.01 


3.53 


Far West 


.48 


.91 


.79 


.43 


.16 


Unallocated 


— 


.01 


— 


.03 


.01 



Source: Mineral Resources, 1921, Part II, Back Folder; Mineral Resources, 1931, Part II, 
p. 430; Minerals Yearbook, 1951, p. 312. 



of the comparatively small capital involved. Not before 1908, however, did 
annual open-hearth output surpass the Bessemer but by the early 1950's; the 
Bessemer process had declined to about 3% of all steel output. 

A third major process, the electro-metallurgical technique, was first intro- 
duced on a large scale in the United States during the winter of 1909-10 at the 
South works of Illinois Steel. World War I demonstrated the superiority of the 
electric technique over the crucible in meeting new demands for alloy-steel 
ingots and castings, and the increasing capacities of electric furnaces, in com- 
mon with most others since that time, has brought their product for some 
purposes into competition with the open-hearth. 32 

The growth of an immense iron and steel making complex in the Northeast 

32 Bessemer production was surpassed in quantity by electro-metallurgical steel during 
World War II. Nevertheless, in the early 1950's, the open-hearth still contributed about 90% 
of all United States Steel production which passed 120,000,000 net tons in 1954; the electric 
furnaces made up only 7%, while Bessemer's remaining 3% has increasingly gone into 
dephosphorized steel which has been found to possess the good welding and free-machining 
qualities of original Bessemer. According to some industry sources the most important 
recent adaptation of Bessemer has been in the production of seamless pipe. For that matter, 
the enlarged market for all steel products during this century has been a consequence of 
improvements in rolling and finishing practices which make steel not only the primary 
metal for all heavy manufacture and construction but also for a wide variety of consumer 
goods formerly made of wood. 

American Iron and Steel Institute, Annual Statistical Report, 1956, p. 9-11. Among 
purely manufacturing categories, iron and steel continued to be the principal consumer of 
capital during most of this century, although it was only fifth among all categories. After 
1900 most new capital in iron and steel went into the conversion from Bessemer to open- 
hearth in order to take fuller advantage of Superior ores. After 1920 the continuous strip 
rolling- mill producing thin standard sheets for the automobile industry became the largest 
capital-consuming innovation; T. C. Cochran, op. cit., pp. 30, 45-6. 



214 / Regional Economic Development: 1870-1950 

was clearly rooted historically in ready access to major deposits of coking coal 
and in the availability of local ore, as well as in access to the major markets 
of the nation. The bog iron of the Atlantic seaboard had been used up in the 
eighteenth century and the coal mines of eastern Virginia, whose development 
Hamilton had wanted to stimulate by an import duty, were abandoned soon 
thereafter. Before another century had passed, the iron resources of the Appala- 
chian range had gone into decline and the anthracite coals of eastern Pennsyl- 
vania could be worked only at steadily increasing cost. 33 By the late 1890's a 
pronounced drift of the iron and steel complex into the Great Lakes states was 
evident, following the large-scale development of Minnesota iron, but the force 
of this attraction was diminished by the greater weight of fuel requirements in 
treating Mesabi ores. 34 Within the next two decades, however, advances in fuel 
technology reduced the bondage to fuel supplies, and various market factors had 
begun to exert a more decisive influence. The subsequent drift of steelmaking 
within the Northeast, therefore, reflects a locational force stemming from coun- 
trywide markets, notably the variety of demands for structural steel and the rise 
of the automobile industry. Nevertheless, the major iron and steel markets have 
coincided historically with fuel sites in general and coal sites in particular. 
When coal was the paramount locational factor, most other heavy industry and 
processing trades oriented to iron and steel converged on the coal sites. Because 
so many manufacturing activities were themselves heavy users of coal (either as 
coke or electricity, the classic locational pattern of industrial revolution was set 
in the Northeast. 35 In the absence of primary markets outside the more urban- 

33 P. Roberts, The Anthracite Coal Industry (New York: Macmillan Co., 1901), chap. 
I-IV and XL The waste of coal fuel in the late 19th century was prodigious. The waste 
in mining was computed at from 40 to 60%; steam engines utilized about 8% of the coal 
they burned and not 10% of the fuel burned in power plants was converted into energy. 
Electricity generating stations were said to utilize less than 1% of the coal consumed; K. 
Coman, The Industrial History of the United States (New York: The Macmillan Co., new 
and revised edition, 1918), p. 385. 

34 The effect of Mesabi ore on the coke consumption rate at blast furnaces became clear 
during the first decade of this century. In 1902 when the proportion of Mesabi ore in 
burden was 43.8%, the coke rate (lbs. per net ton of iron) was 2,155; by 1907 when the 
proportion of Mesabi ore had risen to 68.7%, the coke rate was 2,362. This temporary 
reversal of declining fuel consumption was due primarily to an inability to control the 
channeling of gases and general irregular working; H. S. Brassert, "Modern American 
Blast Furnace Practice," in American Iron & Steel Institute, Yearbook, 1914, p. 40. 

35 W. Isard, "Some Locational Factors in the Iron and Steel Industry since the Early 
Nineteenth Century," Journal of Political Economy, Vol. LVI (1948), pp. 203-17. Since 
transfer costs are generally minimized by moving the heaviest goods the shortest distance, 
the mass of heavy industry can still be most efficiently located in terms of the energy 
costs of transporting primary fuels and ore. Under modern conditions, however, the avail- 
ability of cheaper fuels for processing and fabrication by maritime transport or low-cost 
pipeline, and differentials in space-heating costs due to climatic variation, may make a 
geographically distant point more economical than one within a few miles of a coal field. 
The critical factor, therefore, is total energy cost and not simply distance; F. Cottrell, 
Energy and Society; The Relation between Energy, Social Change, and Economic Develop- 
ment (New York: McGraw-Hill Book Co., 1955), pp. 296-98. The enduring strength of 
the Northeastern industrial belt and the selective migration of manufactures to other regions 
since 1910, detailed earlier, provides documentation of this view. 



Nationalizing the American Economy, 1870-1910 / 215 

industrialized areas, the external economies of local concentration combined 
with the immense capital commitment in existing facilities to impose profound 
restraints on any wholesale relocation of heavy industry. 36 



Differentiation in Lumber Production 

For a long period after the Civil War, the almost universal material for con- 
struction and fabrication was not iron and steel, but wood. The homes, mills, 
vehicles, and much of the mechanical equipment and fuel with which the West 
was initially developed had their origins in the great forests. The period 1870- 
1910, therefore, witnessed the peak of lumber production and the beginnings 
and rapid ascent of pulp and paper manufacture; it also produced the first signif- 
icant impact of the conservation movement as the public awoke to the danger of 
forest depletion. 37 Already before 1870, production centers of the lumber indus- 
try were shifting from the New England and Middle Atlantic regions into the 
Great Lakes states, where the commercial practices of the great lumber com- 
panies proved even more destructive than the broad ax of the pioneer. By the 
1890's much of the best timber had been cut away from the upper Mississippi 
Valley (white pine) and heavy demands were being made upon the pine barrens 
of the Gulf states. The forests which had for long sheltered the sugar plantations 
and citrus groves were also being laid low. Before 1910 more than a fifth of the 
great Carolina forests had been removed and structural timbers were being 
shipped the width of the continent from Oregon to Maine. The nation's insati- 
able appetite for lumber, in fact, had compelled migration from the cutover 
sections of the Great Lakes and Plains regions into the Far West, where hitherto 
only timbers close to rivers had been heavily felled. 38 

36 The steel capacities of other regions, even in the age of "foot-loose" industry, are severely 
limited by the size and character of local demand. Before World War II the only fully 
integrated steel plant in the Mountain or Far West regions was at Pueblo, Colorado. Those in 
Utah, California, and the Pacific N. W. were small- and medium-scale mills operating mostly 
on scrap. Western expansion, under government auspices, has raised the capacity of these 
two regions to nearly 6% of the national, but a recent study indicates that the existing 
market in those parts is closer to 10% of the national; C. L. White, "Is the West Making 
the Grade in the Steel Industry?" Business Research Series No. 8 (Stanford: Stanford 
University Graduate School of Business, 1956). 

37 As late as 1940 woodpulp had probably not yet displaced fuel as the second largest 
consumer of the forests, although the peak of fuel-wood consumption was reached around 
1890. Indeed, only after that date did lumber sawed exceed the lumber equivalent burned. 
In the entire eastern half of the country and the Plains, wood burned was probably a 
greater commodity drain than the entire product of the saw mills. Much of this wood, of 
course, was not suited to lumber use and the drain was in any case much reduced before 
the assault began upon the softwood forests of the interior and the Pacific N. W.; R. V. 
Reynolds and A. H. Pierson, "Fuel Wood Used in the United States, 1630-1930," U.S.DA. 
Circular No. 641. (February 1942) , pp. 14-15. 

38 U. S. National Conservation Commission, Report (Washington: Government Printing 
Office, 1909), Vol. 1, pp. 51-73; Vol. 2, pp. 179-269, 547-80, 498-511, and 748-58. The 
Commission reported that since 1900 the cost of white pine lumber had increased 53%, oak 
54%, hemlock 56%, Douglas fir 63%, yellow pine 65%, and poplar 78%. About two-thirds 
of timber felled never reached the market. 



216 / Regional Economic Development: 1870-1950 

In 1910, therefore, the forest economy of the United States was a nationwide 
organization with major sources of supply located far from the chief centers of 
consumption. It is testimony once more to the thrust of technology that the 
period of the lumber industry's greatest efflorescence was also the final stage in 
the transition to the age of iron and steel. 

The "free land" that drew the covered wagons westward and helped build 
the transcontinental railroad was no less attractive to the lumberman than to the 
farmer or the dealer in railway real-estate. The forest story is an integral part 
of the epic of the public lands. 39 The westward movement of the lumbermen 
after the destruction of the Appalachian forest cover was initially not unlike 
the parallel migration of the farmers. Unlike mining development, the forests 
were not at that period dominated by a handful of great corporations but by "a 
profusion of partnerships and coalitions of partnerships amid a general environ- 
ment of intense competition." 40 Nevertheless, before the close of the century the 
lumber industry had also responded to the institutional tendencies of the day, 
increasing its investment, expanding its facilities, and organizing against frantic 
competition. Regional and nationwide trade associations had emerged, some- 
times in rivalry, as when Great Lakes groups tried to close the market to 
Southern pine, and at other times in co-operation, as in efforts to stabilize 
prices. 41 The main impulse to expansion and association had its roots in the 
problem of securing countrywide distribution from regional production centers 
and in the need to concentrate large holdings of virgin timber in a few hands in 
order to forestall competitive cutting. Important, if not spectacular, changes in 
technique slowly reinforced these market pressures. 42 

The development of techniques for making paper from woodpulp had pro- 

39 0f the four principal land acts passed during the 1870's, at least three were directly 
concerned with Western timberlands and met the needs of special interests operating there: 
the Timber Culture Act of 1873, the Timber and Stone Act of 1878, and the Timber 
Cutting Act of the same year. These measures, subject to flagrant abuse, remained in 
force until Cleveland's General Revision Act of 1891 which remedied some of the worst 
defects in public lands disposition and authorized the establishment of forest preserves, the 
first step in the national conservation movement; R. G. Lillard, The Great Forest (New 
York: Alfred A. Knopf, 1947), pp. 233-76; R. K. Winters, ed., Fifty Years of Forestry in 
the U.S.A. (Washington, D. C; Society of American Foresters, 1950), pp. 1-29; R. F. 
Fries, Empire in Pine: The Story of Lumbering in Wisconsin, 1830-1900 (Madison: State 
Historical Society of Wisconsin, 1951), pp. 161-203; W. A. Rowlands, "The Great Lakes 
Cutover Region," in M. Jensen, ed., Regionalism in America, pp. 331-46. 

40 F. W. Kohlmeyer, "Northern Pine Lumbermen: A Study in Origins and Migrations," 
Journal of Economic History, Vol. 16 (December, 1956), pp. 529-38. 

41 R. C. Bryant, Lumber, Its Manufacture and Distribution (2nd ed., New York: John 
Wiley and Sons, 1938), pp. 309-45; W. B. Greeley, Forestry and Men (Garden City, N. Y.: 
Doubleday and Co., 1951), pp. 30-63. 

42 P. L. Buttrick, Forest Economics and Finance (New York: John Wiley and Sons, Inc., 
1943) , pp. 217-18, lists the following among principal advances in logging and saw mill 
practice before World War I: Construction of railways deep into the woods permitted the 
cutting of timber formerly too remote from streams for river driving; this facilitated year- 
round logging and the movement of hardwoods not buoyant when green. Powered loaders 
and skidders cheapened and accelerated the removal of logs from forests to the mills. 
Within the mills, band-saws and other automatic machinery increased labor productivity. 



Nationalizing the American Economy, 1870-1910 / 217 

found repercussions on the regional distribution of forest activities. The Tilgh- 
man experiments with effects of various acids on wood had commenced about 
1867; sulphurous acid was found to dissolve the ligneous elements readily and 
to leave a residue of cellulose fibers. Improvements in this first sulphite process 
were carried out in Europe and incorporated into the first American pulping 
plants before 1880. A variety of other processes were perfected before the end 
of the century, the most significant of which were the sulphate (1884) and 
caustic soda methods. For most types of paper only the cellulose fibers can be 
employed but in the production of newsprint almost the entire fibrous matter can 
be used and hence a cheaper, purely mechanical, grinding process will suffice 
to convert wood into pulp, provided small quantities of sulphite pulp are added 
to give the paper body. 43 By no means all species of wood are adaptable for 
pulping, and for a long time only spruce and hemlock could be utilized. Small 
supplies of these species were available near some of the New England and New 
York paper mills which were already oriented to the existing source of materials 
(straw, rags, waste paper, and manila stock) readily obtainable near the 
industry's larger urban markets. But as late as 1910 the larger and apparently 
more viable paper mills were still locating in areas where pulpwood was either 
scarce or non-existent; by this time the most accessible forests where the desired 
species were still abundant were situated in Maine and Wisconsin. 44 

Exhaustion of the remaining pulpwood stands in the New England and Mid- 
dle Atlantic states before World War I was accelerated by a brief revival of 
lumbering in these regions. The shift of lumbering into the Plains and western 
Great Lakes states three or four decades earlier had tended to raise the price 
of lumber in leading Eastern markets and, as a consequence of the construction 
boom after 1900, it became economical once more to start cutting in previously 
inaccessible reaches of the Eastern forests. This belated invasion, moreover, was 
facilitated by the recent introduction of compact and highly mobile saw mills. 
But from the standpoint of the pulp mills, this last spasm of lumbering proved 
an unmitigated calamity. Since the preferred white pine had already been cut 
out, the main weight of the new assault fell on the spruce and hemlock, leaving 
the pulp industry dependent for supplies on wood from abroad or from other 
more remote regions of the United States. 45 Paper users, especially the news- 
print industry, immediately sought and obtained some reduction of tariffs; as a 
result, after 1913 the position of Eastern newsprint producers was further under- 
mined by rising imports from Canada. 46 

43 The quantity of fibrous material which needs to be separated from wood constituents 
varies with the end product sought. The U. S. Tariff Commission, Report to the United 
States Senate on Wood Pulp and Pulpwood, Report No. 126, 2nd Series (Washington: 
Government Printing Office, 1938), pp. 38-43, gives a clear account of pulping processes; 
also C. R. Hall, op. cit., pp. 274-79. 

44 More than 50 of the 135 pulp mills in operation or under construction in 1882 had 
closed by 1899; H. Hunter, "Innovation, Competition, and Locational Changes in the Pulp 
and Paper Industry: 1880-1950," Land Economics, Vol. 31 (1955), pp. 317-18. 

K Ibid., pp. 318-19. 

46 J. A. Guthrie, The Newsprint Paper Industry, An Economic Analysis (Cambridge: 
Harvard University Press, 1941). 



218 / Regional Economic Development: 1870-1950 

The force of foreign competition was felt even in the Great Lakes region, and 
the locus of new pulping activity shifted closer to its wood supply in the Pacific 
Northwest or the South. Meanwhile, surviving paper manufacturers on the 
Eastern seaboard could stay in business only by closing their newsprint and low- 
grade papermaking plants and concentrating their facilities on high-grade 
papers, such as book stock and special printing, writing, and tissue paper. After 
World War I it became more economical to ship the processed woodpulp rather 
than the raw pulpwood and the possibility of cheap water transport via the 
newly opened Panama Canal now gave the stranded paper industry of the North- 
east a further lease on life. Improvements in the sulphate process proved a 
special boom to makers of crude wrapping or "Kraft" papers in the South; 
owing to the belated adoption of improved forestry practices, producers here 
could henceforth compete favorably with their rivals in the Pacific Northwest. 
This heightening of interregional competition gradually forced the paper indus- 
try as a whole to move closer to the sources of pulp in the Southeast and Far 
West, while down to World War II at least, the more expensive high-grade 
papers maintained their hold in the older Northeast. 47 



Mass Production for National Markets 

By the opening decade of the twentieth century the economy of the United 
States was effectively nationalized. Both resource and processing activities were 
slowly and selectively adjusting to the conditions of a nationwide market. The 
railroad network extended to every part of the continental territory and in some 
regions was supplemented by a somewhat dilapidated system of inland water- 
ways. Coastal shipping had recently witnessed a notable revival, and the early 
efforts of Standard Oil to prevent the growth of a system of oil pipelines in 
competition with their captive railroads had failed. 48 Already the promise of 
the automobile was stimulating interest in the long-neglected public highways 

47 By 1930 more than half the pulpwood used in Pennsylvania and New York was im- 
ported from Canada and Sweden and by 1940 the proportion of imported wood used in the 
paper states of the Lakes region, Michigan and Wisconsin, approximated one-quarter; H. 
Hunter, loc. cit., pp. 322-23; J. A. Guthrie, The Economics of Pulp and Paper (Pullman, 
Wash: The State College of Washington Press, 1950), pp. 81, 143-64. 

48 G. R. Taylor and I. D. Neu, The American Railroad Network, 1861-1890 (Cambridge: 
Harvard University Press, 1956) ; A. M. Johnson, The Development of American Petroleum 
Pipelines: A Study in Private Enterprise and Public Policy, 1862-1906 (Ithaca, N. Y.: 
Cornell University Press for The American Historical Association, 1956) ; Inland Water- 
ways Commission, Preliminary Report (Senate Document No. 325, 60th Congress, 1st 
Session [Washington: Government Printing Office, 19081), pp. 177-312; U. S. Commissioner 
of Corporations, Transportation by Water in the United States: Report (Washington: 
Government Printing Office, 1909), Vol. I, pp. 149-380 and Vol. II, pp. 249-80. The ex- 
tension of coastwise shipping regulations to Alaska, Hawaii, and the Philippines helped 
raise the Pacific registered tonnage 120%, 1897-1909. Total coastwise tonnage, including 
Lakes and Western river vessels, in 1909 was 6,500,000 or eight times the tonnage registered 
for ocean trade. Also, H. U. Faulkner, "The Decline of Laissez-Faire, 1897-1917" in The 
Economic History of the United States (New York: Rinehart and Co., 1951), Vol 7, pp. 
220-48. 



Nationalizing the American Economy, 1870-1910 / 219 

and would shortly result in a burst of construction in and between the major 
metropolitan centers. Of course, spatial frictions had not been eliminated; but 
all parts of the country were now accessible — at a price. 

Institutional or "artificial" distances had acquired a greater economic signifi- 
cance than the actual physical distances they were designed to reflect. After 
1887 the Interstate Commerce Commission had given legal sanction to railway- 
association rates and classifications which, in effect, had divided the country into 
four or five "rate territories". There was a so-called "Official Territory" that 
corresponded almost exactly with the greater Northeastern industrial belt; its 
boundaries ran along the Mississippi River in the west, then east along the Ohio 
and Potomac, and finally southeast to the Atlantic coast in Virginia. There was 
also a "Southern Territory," which included the states that comprise our South- 
east except Virginia, Arkansas, and Louisiana, and a "Southwestern Territory" 
made up of Arkansas, Louisiana, Texas, and parts of New Mexico. The rest of 
the greater West was eventually divided into two further territories. Low rates 
for raw materials moving into the "Official Territory" and high rates for manu- 
factured items produced outside it undoubtedly discouraged and retarded the 
diffusion of manufactures into outlying regions where they might have derived 
some cost advantage from proximity to their raw materials. 49 Undoubtedly, 
throughout this period, the "natural" and "artificial" costs of moving heavy 
freight were a formidable "imperfection" in the working of the American 
market system. But by 1910, increasing competition among the different trans- 
port media, together with more effective public regulation, was beginning to 
effect a greater equality of net per ton mile rates. 50 

The reduction of interregional freight differentials at this time paralleled a 
convergence in the levels of per capita personal income in the various regions. 
Differences in personal income were relatively greater before 1880 than at the 
end of World War I. While the rich mineral and forest resources of the Moun- 
tain and Far West regions helped to raise the personal income levels of their 

49 For a description of freight traffic movements by class of commodity originating on the 
lines of reporting railroads, see Industrial Commission, Vol. 19, Final Report, op. cit., pp. 
264-66. The consequences of discrimination were later described as follows: "For nearly 
three generations . . . the country has been accepting influences and conditions which have 
made a workshop region of one section . . . , with the other sections, as marked off by the 
present territorial rate boundaries, being largely in a position of contributors of raw 
materials and semifinished products to the industrialized region." J. H. Alldredge, The 
Interterritorlal Freight Rate Problem of the United States, House Document No. 264, 75th 
Congress, 1st Session (Washington: Government Printing Office, 1937), p. 51. 

50 But as recently as 1938 it was estimated that, on average, shippers of classified goods in 
the Official Territory paid 75% less for the same railroad services than shippers in the 
Southwestern Territory, 71% less than those in the Western Trunkline Territory, and 35% 
less than shippers in the Southern Territory. Five years later these differentials had been 
still further reduced, but it was affirmed in a Report on Interterritorial Freight Rates that 
"territorial differences in the levels of first-class rates . . . are large, but the trend during 
the past 25 or 30 years has been towards reducing the spreads." House Document No. 303, 
78th Congress, 1st Session, pp. 21 and 55-57. Also, Regional Freight Rates: Barrier to 
National Productiveness, ibid., No. 137. By no means all of these differentials, of course, 
can be attributed to "discriminations." 



220 / Regional Economic Development: 1870-1950 

small populations far above the national average, the rapidly industrializing 
Middle Atlantic and Great Lakes regions had also achieved a substantial ad- 
vance. For a period, only the war-devastated Southeast appeared unaffected by 
the surge of national income. But as population continued to pour into the 
regions of the newer West, per capita incomes everywhere settled closer to the 
national average. This tendency toward convergence did not, of course, oblit- 
erate sizable absolute differences, but it did bring about a situation where 
families in every section of the country were extending the range of goods and 
services they wanted and could afford to acquire. Even farm families, except in 
the South, were enjoying unprecedented prosperity and could expect much the 
same attention from manufacturers and distributors as their neighbors in the 
city. The United States had, in fact, become a vast national "free trade" 
market. 51 

Unlike the national markets of many smaller European countries, the Ameri- 
can market was essentially a mass market. It demanded and obtained large 
quantities of cheap low-to-medium quality goods with comparatively little varia- 
tion from one part of the country to another. Where products could be stand- 
ardized, as was already evident in a few instances at the time of the Civil War, 
the logic of growth required that production be mechanized. 52 The essential 
condition for an economical system of mass production was its complement, a 
mass market; this condition and a number of other preconditions had already 
been met by 1910. 53 In volume, the production of durable goods for capital 
equipment rose at an average annual rate of 5% between 1901 and 1913 as 
against 2.6% for goods intended for human consumption, though in value terms 
the capital goods sector probably amounted to not more than 20% of total goods 
production even in the peak years. Significantly, this great surge in industrial 
production, which according to one index more than doubled over the years 
1899-1919, "was due largely to the development of our mineral resources and to 
the increasing volume of fabricated goods into which raw mineral products 
enter." 54 

Hardly less significant for the shaping of the economy than the growth of 
mass production and marketing was the fact that most of the primary sources 
of materials — the mines, fields, and forests — were located within the same ter- 

51 G. B. Hotchkiss, Milestones of Marketing (New York: The Macmillan Co., 1938), p. 220. 

52 The shortage of skilled labor outside of the mechanical trades and the ability to import 
European products on increasingly favorable terms (subject to the tax of the tariff) meant 
that the relatively small American "quality" markets could be supplied by imports or by 
the small-scale operations conducted by immigrants steeped in the tradition of old-world 
craftsmanship. 

53 These preconditions include the capacity for making very fine heavy-duty machine tools 
(which had recently been enhanced by the perfection of a new high-speed carbon steel 
during the years 1900-06 by F. W. Taylor and J. M. White), as well as the ability to 
mobilize adequate capital through a specialized capital market over and beyond the usual 
profits plowed back into expansion. 

M F. C. Mills, Economic Tendencies in the United States (New York: National Bureau of 
Economic Research, 1932), pp. 13, 21-23. According to Mills the average annual rate of 
increase in the physical volume of non-durable goods (raw and processed) production was 
2.5% over the years 1901-13, that of semidurable (mostly textile and leather) production 



Nationalizing the American Economy, 1870-1910 / 221 

ritorial-political boundaries. Unlike the manufacturing nations of Europe, the 
economy of the United States was for most material and market purposes "self- 
sufficient." Whereas England and Germany must bring increasingly large 
volumes of materials from abroad, "the industrial progress of the United States 
was the result of carrying labor to raw materials." 55 Though abundant, these 
materials were irregularly distributed throughout the continental territory and 
hence the comparatively unskilled, if highly mechanized, labor force had to be 
able and willing to move on to primitive conditions of successive frontiers. The 
existence of these unused resources did not, however, guarantee industrial prog- 
ress. Only by harnessing the power of machines and by deploying capital in 
corporate array could the outlying regions be organized to render up their 
growth-sustaining material to the corpus of national economy. This involved a 
restructuring of the labor force among extractive, processing, and servicing 
activities, and the redistribution of these activities among regions across the 
continental territory. This marks the real distinction between the earlier period 
of loose-knit economic sectionalism and the later period of integrated national 
economy. Whereas the pioneer in the ante-bellum West was on the frontier of 
an essentially mercantile-colonial economy, his children and grandchildren, 
whether farmers, miners, loggers, or urban workers, were increasingly full par- 
ticipants in the industrialization of the entire nation. It was the dynamic inter- 
play of technology on the natural resources of both the more and the less devel- 
oped regions, joined with human ingenuity and industrial organization, that in 
the period 1870-1910 brought about a continental system of exchange and. with 
it, the most affluent level of living in recorded history. 



2.6%, and that of durable goods production 4.6%. The industrial production index referred 
to is that of Day and Thomas which stood at 100 in 1899, 159 in 1909, and 214 in 1919; 
E. E. Day and W. Thomas, The Growth of Manufactures, 1899-1923 (Census Monographs 
VII, 1928), Washington, D. C: Government Printing Office, pp. 32-4. 

55 V. S. Clark, History of Manufactures in the United States (3 Vols., New York: McGraw- 
Hill Book Co., 1929), Vol. II, p. 2. By 1900 American per capita consumption of manu- 
factured goods was almost half as much again as that of Britain and double that of 
Germany and France; S. Rezneck, "Mass Production since the War between the States," 
in H. F. Williamson, ed., The Growth of American Economy (New York: Prentice-Hall, 
Inc., 1946), p. 498. 



15 / Regional adjustments within the 
national economy, 1910-1950: 

(1) POPULATION SHIFTS 



In the first half of the nineteenth century the urban-industrial form of economic 
development was a highly localized phenomenon grounded in the original pat- 
terns of settlement and local advantage obtained under the technology-resource 
potentials of the time. During the second half of the century urban-industrial- 
ization became generalized as resource depletion in the more advanced areas, the 
rapid progress of technology, and the growth and migration of population neces- 
sitated both the diversification and extension of the economy's material base. 
By the first decade of the twentieth century, some narrowing of the differentials 
among regions in industrial structure and in levels of living was apparent in 
spite of different conditions and different stages of development. In this sense, 
the American economy was effectively nationalized. Subsequent growth and 
adjustment turned on further refinement in the system of interdependence 
among the regions for the greater benefit of the whole. 

By 1910 the dominant regional patterns with which we are today familiar had 
already taken shape. Thus, for example, in spite of many technical advances in 
the important light-metals and alloy-steel industries and shifts in the composition 
of the industrial fuel supply, manufacturing is still almost as much the spe- 
cialized concern of the Northeastern industrial belt as it was in 1910. Yet the 
apparent fixity of the twentieth century regional structure can be overstated. It 
is no small matter that petroleum reginning and related industries, certain 
phases of the lumber products industry, and more recently the aircraft and 
electronics industries have taken root in parts of the Southeast, Southwest, and 
Far West regions rather than in the Northeast. Nor can the fact be lightly dis- 
missed that, within the Northeastern industrial belt itself, the bulk of growth in 
the twentieth century has taken place in the Great Lakes region rather than in 
the Middle Atlantic or New England states. There has, moreover, been some 
general diffusion in manufacturing activity in close association with movements 
of population. But far more striking than these changes have been shifts in 
resource activity and concomitant regional growth. 

222 / 



Regional Adjustments within National Economy, 1910-1950 (1) / 223 

Developments in the twentieth century, it is clear, have been characterized far 
less by the gross, nation-building sweeps of movement than those of the previous 
century. Throughout the nation, regional economies have become more com- 
plex, and significant internal (intraregional) changes as well as powerful exo- 
genous changes have to be reckoned with. Within the national framework, for 
example, growth has been increasingly affected by the drawing force of regional 
markets and changing concentrations of populations within metropolitan areas. 

Because of the growing complexity and refinements of the sub-national econ- 
omy, a full understanding of regional economic developments of the recent past 
calls for analysis in substantial detail. This is undertaken in Parts IV and V of 
this book. Here, we ishall sketch only in broad outline the changes in the 
regional structure of the U. S. economy after 1910 as revealed by the same key 
indices that we have employed in describing the period 1870-1910, that is, 
regional changes in population, productive activities, and income. 



Growth and Shift in Population, 1910-1950 

Between 1910 and 1950 the rate of population growth in the United States 
slackened: in the 40-year period from 1870 to 1910, population numbers had 
risen by more than 130% ; in the ensuing 40-year period they rose by less 
than 64%. 



Table 79. Percentage increases in population by census decades, by region, 
1910-1950 



Region 


1910-20 


1920-30 


1930-40 


1940-50 


1910-50 


United States 


14.9 


16.1 


7.2 


14.5 


63.8 


New England 


12.9 


10.3 


3.3 


10.4 


42.1 


Middle Atlantic 


15.3 


17.4 


5.8 


11.0 


59.0 


Great Lakes 


17.7 


17.8 


5.3 


14.2 


66.6 


Southeast 


10.5 


12.2 


10.6 


12.0 


53.5 


Plains 


7.8 


6.0 


1.7 


4.0 


20.8 


Southwest 


21.4 


22.9 


7.7 


16.3 


86.9 


Mountain 


26.9 


7.3 


9.3 


15.8 


72.5 


Far West 


32.1 


46.8 


18.8 


48.8 


242.7 



Figures in italics are above the national average. 
Source: 17th U. S. Census 1950, see supra, Table 1, note. 



Regional populations generally reflected the broad inclinations of the national 
movements: prosperity, depression, and war are mirrored in the decennial 
fluctuations of population growth. After World War I, however, the trend in 
most regions was in closer conformity with the national trend than it had been 
before 1910, and only the Far Western states bordering the Pacific experienced 



224 / Regional Economic Development: 1870-1950 

rapid growth comparable with that in the earlier period (Table 79). By 1930 
the Southwest had already fallen into closer step with countrywide development, 
while the surge of population into the Mountain region had ended a decade 
before. Among Western regions, however, only the Plains failed to maintain 
the national rate of growth; it continued in a phase of relative decline which had 
set in after 1890. By 1930, in fact, Minnesota, Nebraska, and Kansas (like Iowa 
and Missouri earlier in the century) had fewer in-migrants than out-migrants, 
and between 1930 and 1940 the western Plains states experienced an absolute 
decrease in numbers. 

The urban-industrial regions of the Northeast reveal some significant con- 
trasts with the earlier period. New England fails to maintain the national rate 
of population growth but, for a while, the Middle Atlantic and Great Lakes 
regions are the beneficiaries of an "eastward" movement. 1 The Middle Atlantic 
region had, to be sure, been growing at a faster rate than the nation since 1890, 
but after 1910 the Great Lakes region had a comparable rise; both regions con- 



Table 80. Density of population, by region, 1910, 1930, and 1950 

(population per square mile of land area) 



Region 


Area* 


1910 


1930 


1950 


United States 


2,974,726 
sq. mi. 


30.9 


41.3 


50.7 


New England 


( 2.12) 


103.7 


129.3 


147.5 


Middle Atlantic 


( 3.78) 


188.1 


254.5 


299.1 


Great Lakes 


( 8.23> 


74.5 


103.3 


124.1 


Southeast 


(17.95) 


41.2 


51.1 


63.3 


Plains 


(17.17) 


22.8 


26.0 


27.5 


Southwest 


(19.08) 


10.7 


16.0 


20.0 


Mountain 


(17.23) 


3.9 


5.4 


6.8 


Far West 


(14.44) 


9.9 


19.3 


34.1 



Figures in italics are above the national average. 

* Figures in parentheses show proportion of total U. S. land area 1950. 

Source: Statistical Abstract of the United States, 1954, Table 4, p. 10; 17th U. S. Census, 
1950; see supra, Table 1, note. 



tinued in the ascendent until the onset of depression in 1930. Only after the 
outbreak of World War II was there any notable recovery in the general west- 
ward movement, and it was somewhat more selectively directed now towards 
parts of the Far West, Southwest, and southern Mountain states. Throughout the 
years between 1870 and 1930 the Southeast lost relatively, but between 1930 and 

1 C. Goodrich, et al., Migration and Economic Opportunity (Philadelphia: University of 
Pennsylvania Press, 1936), pp. 675-83; George Soule, "Prosperity Decade; From War to 
Depression, 1917-29" in The Economic History of the United States (New York: Rinehart 
and Co., 1947), Vol. 8, pp. 208-15. 



Regional Adjustments within National Economy, 1910-1950 (1) / 225 

1940 the situation was reversed. In that decade its population grew faster than 
that of the nation and relatively faster than that of any other region except the 
Far West. The effect of these disparate rates of growth on regional densities of 
population are summarized in Table 80. 

The effect of the differential rates of growth on regional shares of national 
population after 1910 is clearly much slighter than in the years between 1870 
and 1910 (Table 81). The combined share of the four Eastern regions (New 
England, Middle Atlantic, Great Lakes, and Southeast) in 1870 was 85.5%; in 
1910 it was 73.8% and by 1950 it had declined only to 71.1%. New England, 
of course, has continued in slow relative decline over the whole of the period 



Table 81. Regional distribution of U. S. population, 1910-1950 



(per cent of total) 



Region 


1910 


1920 


1930 


1940 


1950 


United States 


91,972,266 


105,710,620 


122,775,046 


131,669,275 


150,697,361 




100% 


100% 


100% 


100% 


100% 


New England 


7.1 


7.0 


6.7 


6.4 


6.2 


Middle Atlantic 


23.0 


23.1 


23.3 


23.0 


22.3 


Great Lakes 


19.8 


20.3 


20.6 


20.2 


20.2 


Southeast 


23.9 


23.0 


22.2 


22.9 


22.4 


Plains 


12.7 


11.9 


10.8 


10.3 


9.3 


Southwest 


6.6 


7.0 


7.4 


7.4 


7.6 


Mountain 


2.2 


2.4 


2.2 


2.3 


2.3 


Far West 


4.7 


5.3 


6.8 


7.5 


9.7 



Source: 17th U. S. Census, 1950, see supra, Table 1, note. 



Table 82. Principal net shifts in population among states, 1910-1950 





Percentaj 


ge net 




Percentage 


net 






upward 


shift 




downward 


shift 




California 






44.72 


Pennsylvania 




13.78 


Michigan 






11.81 


Missouri 




9.64 


Florida 






10.28 


Iowa 




6.85 


Texas 






8.87 


Virginia 




6.40 


New Jersey 






4.53 


Kansas 




5.78 


Washington 






3.39 


Georgia 




5.55 


Maryland-D. 


C. 




3.21 


Massachusetts 




5.52 


N. Carolina 






2.98 


Kentucky 




5.40 


Oregon 






2.80 


Mississippi 




5.12 


Arizona 






2.77 


Arkansas 




4.48 



Source: 17th U. S. Census 1950, see supra, Table 1, note. 



226 / Regional Economic Development: 1870-1950 

since 1870, but the relative stability of the other two Northeastern regions, since 
1890 at least, is quite remarkable. Even the Southeast has lost relatively little in 
respect of its population share since 1910. 

The growth of the West is merely the reverse of the decline in the East, but 
not all regions shared equally in this growth. The Plains region has been in 
relative decline since the 1890's, and the Mountain region has barely held its 
own since 1920. The Southwest, on the other hand, raised its share of total 
population 13% between 1910 and 1950, and the Far West by a spectacular 
106%. 

Table 82 and Figure 33 show the results of our shift analysis of net upward 
and downward movement across state lines (the extent to which states exceed 
or fall short of the percentage national growth of population) between 1910 and 
1950. The most startling development of this period took place in California, 
which accounted for 44% of the total net upward shift. Both Washington and 
Oregon, among Far Western states, also registered considerable relative growth. 




Absolute upward shift 
j-X-1 Relative downward shift 



Figure 33. Net Shift in Total Population, 1910-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 25.5. Total net shift as % of 1910 population: 16.3. 



In the Southwest, growth after 1910 was chiefly confined to Texas, although 
Arizona also contributed sizably to the region's advance. Oklahoma, however, 
experienced a relative net downward shift of more than 3%, in contrast to its 
relative growth in years before 1910. All of the Plains states contributed to the 
total net downward shift; Missouri, Iowa, Kansas, and Nebraska were among 
the greatest net losers of population, with a combined relative net downward 
shift of 26%. 



Regional Adjustments within National Economy, 1910-1950 (1) / 227 

In the eastern half of the nation, eight states increased relative to the national 
average, almost offsetting the total relative downward shift of the remaining 
states. In the Great Lakes region, Michigan's relative net influx of almost 12% 
compensated for relative declines in Illinois, Indiana, and Wisconsin; as a 
result, this region held its own over-all. Maryland and New Jersey contributed 
combined upward shifts of more than 6% to the Middle Atlantic region but 
failed to offset the marked net downward shifts in Pennsylvania ; New York, how- 
ever, almost held its own. Among the New England states, Connecticut's 1.21% 
of net upward shift was the only symptom of relative growth; other states 
declined quite heavily; Massachusetts registered a net downward shift of more 
than 5%. Three states in the Southeast — Florida, North Carolina, and West 
Virginia — had relative net upward shifts, but only the first two did much to 
counter the effect of sizable downward shifts from the balance of the region. 

One further feature of demographic changes which deserves special attention is 
the continued growth of cities. Table 83 shows the urban proportions of 
regional population in 1910 and in 1950. Every region except New England and 
the Middle Atlantic states experienced considerable urbanization after 1910 



Table 83. Urban population as percentage of total population, by region, 1910 
and 1950 

Region 1910 1950* 

United States 45.7 59.0 



New England 73.3 74.3 

Middle Atlantic . 70.2 74.0 

Great Lakes 52.7 65.7 

Southeast 19.4 38.0 

Plains 33.2 49.9 

Southwest 22.5 55.5 

Mountain 40.7 51.8 

Far West 56.0 62.7 

Figures in italics are above the national average. 
*01d Census definition. 

Source: 17th U. S. Census, 1950, see supra, Table 4, note. 



(Table 84), but only in the Southwest and Southeast was the over-all increase 
greater in the later period than in the earlier one. 2 Since 1920, however, not 
only has the rate of urbanization declined in many parts of the country, but the 
character of urban settlement has changed. Population generally has become 
more urban, but it has been shifting out from the congested older city centers 

Nevertheless the trend had everywhere gone so far by 1950 that the Census Bureau's 
definition of "urban" was found to understate the true size of city populations and a new 
definition was introduced. 



228 / Regional Economic Development: 1870-1950 

into the suburbs and metropolitan areas around. If country people continued to 
drift in towards the major centers of population, city people themselves have 
spread out into the perimeters of urban areas or to small satellite communities 
in their vicinity. 3 The most rapidly growing metropolitan areas since 1910 
have been in the west; the lowest rates of growth in the Northeast. According 
to Amos Hawley's findings, metropolitan areas in the West have had the most 
widely scattered urban population in every census year since 1910, and decon- 
centration has been endemic there since the beginning of this century. Decon- 
centration from city centers into surrounding areas did not become the 
dominant trend until after 1920 in the Northeast and after 1930 in the 
Southeast. 4 

As of 1950, the Far West had moved into the fourth position in the nation 
with regard to proportion of total U. S. urban population, while the limited 
population growth of New England had reduced this region to sixth place as 
compared to third place in 1910 (Table 85). 

The results of applying "shift" analysis to urban population growth among 
the states, 1910-50, are summarized in Table 86 (see also Figure 34). Once 
more California is the leading contributor of new urban growth, accounting for 
more than 30% of the entire upward shift in the years since 1910. New upward 
shifts in Texas, Michigan, and Florida also parallel the net upward movements 




Figure 34. Net Shift in Urban Population, 1910-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 28.1. Total net shift as % of 1910 urban population: 31.3. 

Source : Table 4, note. 



3 A. H. Hawley, The Changing Shape of Metropolitan America: 
1920 (Glencoe, 111.: Free Press, 1956), pp. 12-23. 
Hbid., pp. 146-60. 



Deconcentration Since 



Regional Adjustments within National Economy, 1910-1950 (1) / 229 



Table 84. Decennial rates of increase of urban population, by regions, 1910- 
1950 



Region 


1910-20 


1920-30 


1930-40 


1940-50 


United States 


29.0 


27.3 


7.9 


19.5 


New England 


17.0 


12.3 


1.7 


7.8 


Middle Atlantic 


22.7 


20.7 


4.8 


8.0 


Great Lakes 


35.7 


28.7 


3.9 


14.5 


Southeast 


36.0 


39.3 


18.6 


33.5 


Plains 


22.2 


17.6 


7.9 


17.1 


Southwest 


63.6 


55.0 


19.4 


52.3 


Mountain 


23.8 


14.7 


18.1 


31.0 


Far West 


45.1 


60.3 


14.9 


43.6 



Figures in italics are above the national average. 
Source: 17th U. S. Census, 1950, see supra, Table 4, note> 



Table 85. Rank order of regions by rate of increase in urban population, 
1900-1910 and 1940-1950, and regional proportions of U. S. urban population, 
1910 and 1950 





1900-10 


1910 


1940-50 


1950 


Region 


% rate 


proportion 


% rate 


proportion 


(1) Southwest 


118.3 


3.3 (7) 


52.3 (1) 


7.1 (7) 


(2) Far West 


112.1 


5.7 (6) 


43.6 (2) 


10.3 (4) 


(3) Mountain 


68.6 


2.0 (8) 


31.0 (4) 


2.0 (8) 


(4) Southeast 


48.3 


10.2 (4) 


33.5 (3) 


14.4 (3) 


(5) Middle Atlantic 


34.6 


35.3 (1) 


8.0 (7) 


28.0 (1) 


(6) Great Lakes 


33.2 


22.9 (2) 


14.5 (6) 


22.5 (2) 


(7) Plains 


31.3 


9.2 (5) 


17.1 (5) 


7.9 (5) 


(8) New England 


25.3 


11.4 (3) 


7.8 (8) 


7.8 (6) 



Rank order in parentheses. 
Source: 17th U. S. Census, 1950, see supra, Table 4, note. 



of all population (see Table 82), though urbanization is proceeding faster in 
Texas than in either of the two other states. The growth of Detroit and other 
urban areas in southern Michigan has enabled that state to become the most 
rapidly urbanizing area east of the Mississippi. Another important development 
since 1910 is the sharp increase in urbanization in many parts of the Southeast, 
a region of relatively declining population. Altogether six Southeastern states 
number among the ten most urbanizing states in the nation, 1910-50. whereas 
only two appear in the list of ten achieving greatest relative upward shifts in all 



230 / Regional Economic Development: 1870-1950 



Table 86. Principal net shifts in urban population among states, 1910-1950 





Percentage 


net 




Percentage 


: net 






upward 


shift 




downward 


shift 




California 








30.31 


New York 




25.30 


Texas 








19.95 


Pennsylvania 




22.01 


Michigan 








9.79 


Massachusetts 




16.87 


Florida 








8.38 


Illinois 




6.70 


N. Carolina 








4.28 


Missouri 




5.02 


Alabama 








3.37 


Ohio 




2.81 


Oklahoma 








3.28 


Rhode Island 




2.62 


Tennessee 








2.51 


Connecticut 




2.00 


Virginia 








2.48 


New Jersey 




1.95 


Louisiana 








2.37 


Wisconsin 




1.67 



Source : 17th U. S. Census, 1950, see supra, Table 4, note. 



population. 5 (See Table 82.) Over the entire 80-year period since 1870, the rate 
of relative urban decline has been greatest in most highly urbanized parts of 
the country — in the Middle Atlantic and New England and, after 1910, in the 
Great Lakes region (except Michigan) and some of the Plains states. It is to be 
expected that highly urbanized states like New York, Pennsylvania, and Mas- 
sachusetts, Which have long histories of urban-industrial concentration, cannot 
maintain the same percentage rate of city growth as in former times. 

We have already noted that, from the standpoint of size distribution, urban- 
industrial concentration in the more developed regions assumes the pattern of 
"urban hierarchy." By 1910, the New England, Middle Atlantic, and Great 
Lakes regions had developed veritable pyramids of cities, ranging upward from 
a broad base of small centers with populations of from 50,000 to 100,000 to an 
apex of large metropolitan centers with three-quarters of a million or more 
inhabitants. By 1950 the metropolitan pattern of urban hierarchy had become 
even more wide-spread across the country, extending, though in truncated form, 
into the more accessible populated parts of the Southeast and Mountain regions 
(see Table 87). Eleven cities had populations exceeding three-quarters of a 
million in 1950, and every region but the Mountain region had at least one 
center with a half-million or more inhabitants. 

By 1950 almost half of the nation's growing number of standard metropolitan 
areas had begun to merge into some eighteen larger "urban-regions," having a 
total population in excess of 60,000,000. Many cities and their surrounding 
satellite areas, formerly demarcated on the map into separate metropolitan areas, 
now tended to overlap in vast complexes of homes, factories, and commerce. 6 

5 No Southeastern state appears in the list of ten experiencing greatest relative downward 
shifts in urban population, whereas five are listed among the ten registering greatest rela- 
tive downward shifts in all population after 1910 (Table 82) . See R. B. Vance and 
N. J. Demareth, The Urban South (Chapel Hill: University of North Carolina Press, 1954), 
pp. 6-107. 

6 The relaxation of urban concentration since 1910 is chiefly attributable to the centrifugal 
forces of automobile transportation and cheap transmission of electric power. 



Regional Adjustments within National Economy, 1910-1950 (1) / 231 

Table 87. Regional distribution of cities in the United States with 50,000 or 
more inhabitants, by size group, 1950 

(000's population) 

Region 50-100 100-250 250-500 500-1,000 1,000+ 

New England 17 11 1 

Middle Atlantic 23 13 3 4 2 

Great Lakes 31 10 4 3 2 

Southeast 23 15 4 1 

Plains 9 4 3 2 

Southwest 10 6 3 10 

Mountain 2 110 

Far West 11 5 5 11 

Source: Statistical Abstract of the United States, 1955, Table 14, pp. 22-23. 



Six of the eighteen urban-regions each contained more than three million 
inhabitants and two others already exceeded one million. Of these eight giant 
conurbations, five were located in the Northeast and three on the Pacific Coast. 
A single urban-region on the Atlantic coast stretched southward from Haverhill, 
Massachusetts, to the far tip of Fairfax County, Virginia, in an almost unbroken 
succession of cities, towns, and suburbs linked together by expressways and 
turnpikes. In 1950 it had a population of 27,500,000, and an estimate for 1957 
credited it with almost 32,000,000 inhabitants. 7 Only five of the eighteen great 
urban-regions had developed in regions other than the Northeast and Far West: 
two in North Carolina, two in Texas, and one in Missouri (which, in many re- 
spects, was merely the westernmost extremity of the Northeastern industrial 
belt) . But there were 65 standard metropolitan areas outside the Northeast and 
Far West: 36 were located in the Southeast, 13 in the Southwest, 12 in the Plains, 
and 4 in the Mountain region. 

Clearly, the growth of metropolitan areas was the characteristic pattern of 
urban settlement in all parts of the United States at the mid-century and, in the 
more industrialized sections, metropolitan areas were already merging into 
urban-regions. In 1950 the urban-regions might embrace only 45% of the 
nation's standard metropolitan areas, but more significant, they already con- 
tained 71% of all metropolitan population. Such were the main demographic 
patterns exhibited in the United States at the close of our period. The westward 
movement of population was now confined largely to the Far West and selected 
localities in the Southwest. The great agricultural heartland of the interior was 
losing population relatively both to the Far West and to the urban-regions of 
the Northeast. Rural population generally was still declining as people moved 
toward the city, but the city itself was rapidly extending its boundaries out into 
the surrounding countryside. In the more developed regions, the Northeast 
and the Far West, the cities were beginning to assimilate one another. 

"New York Times, January 27, 1957, pp. 1 and 72. 



232 / Regional Economic Development: 1870-1950 

The growth of urban-regions also had important implications for the char- 
acter of the American market. The same forces of technology and organization 
that created these vast conurbations have also been directed towards meeting 
the consumption demands of the increasingly urbanized population. The pattern 
of these demands became set during the prosperous twenties, when urban con- 
sumers changed the disposition of their expenditures in the direction of the new 
mass-produced durable goods, processed foods, and services. While there have 
been fluctuations in consumer outlays for durables, non-durables, and services, 
the broad proportions of each have not changed much since the late twenties. 
As a consequence, regional divergences in patterns of personal consumption at 
given levels of income have also been reduced. Such differences as do endure, 
between urban and rural consumers or between rural-farm populations in the 
South and in other parts of the country, for example, are less than they were 
even a quarter of a century ago. 8 It goes without saying, moreover, that the con- 
struction and servicing of urbanized communities during the present century has 
provided a major impetus to economic growth regionally and nationally. Be- 
tween 1900 and 1930 the investment in urban power and light installations alone 
was as large as that of the previous century in railroads. But the strategic 
invention at the heart of the modern metropolitan complex was the automobile. 
The upsurge of automobile production betweeen 1914 and 1929 directly stimu- 
lated the output of oil, steel, rubber, glass, lacquers, and many other items. 
Indirectly, the automobile was responsible for the massive investment in high- 
ways, bridges, service stations, motels, and for the enormous expansion of 
suburbs. 9 It was the style of life emerging in these urbanized regions during 
the second quarter of this century that, on the demand side at least, was shaping 
the character of productive activities in the nation as a whole. 10 



8 U. S. Department of Commerce, Historical and Descriptive Supplement to Economic 
Indicators, 1955 (Washington: Government Printing Office, 1955), pp. 10-11; also M. T. 
Copeland, "Marketing," Recent Economic Changes in the United States (New York: 
McGraw-Hill Book Co., 1929), Vol I, pp. 321-424. On the equalization of per capita dis- 
posable incomes among the states, see C. F. Schwarz and R. E. Graham, Jr., Personal 
Income by States since 1929, a supplement to the Survey of Current Business, Office of 
Business Economics, U. S. Department of Commerce (Washington: Government Printing 
Office, 1956), Table XV, p. 48. 

9 "No one," writes Thomas C. Cochran, "can reliably estimate the vast size of capital 
investment in reshaping society to fit the automobile." The direct and indirect capital 
investment involved "was probably the major factor in the boom of the 1920's." The 
American Business System: A Historical Perspective, 1900-1955 (Cambridge, Mass.: Harvard 
University Press, 1957), pp. 35-45. 

10 The 1954 Census of Business showed that the four superior per capita income regions 
had more than 68% of wholesale trade receipts and almost 64% of retail trade receipts. 
By the early 1950's regional shares of national population coincided almost exactly with 
shares of total receipts from retail trade. In six of the eight regions the correspondence 
was within 2 percentage points. The two exceptions were the Southeast and the Far West, 
the regions of lowest and highest per capita personal incomes respectively. The Southeast's 
share of population was 22.4%, its share of retail trade receipts only 16.6%; the Far West's 
6hare of population was only 9.7%, but its share of retail trade receipts was 12.2%. 
Wholesaling, of course, had a somewhat lower association with population than retailing. 



IQ Regional adjustments within the 

national economy, 1910-1950: 

(2) SHIFTS IN PRODUCTIVE ACTIVITIES 



The main trends in employment structure that were evident before 1910 con- 
tinued unchanged well into the second quarter of the present century. Resource 
activities (agriculture, mining, fisheries, forestry and logging) absorbed a pro- 
gressively smaller share of the working population (Table 88). The manufac- 
turing sector continued to grow through part of the period, but before 1950 its 
share in total employment, like that of the resource sector, shows relative decline. 
Service activities, however, grew rapidly throughout the 1910-1950 period. 



Character of the Labor Force 

In absolute numbers, the labor force grew from some 38 million in 1910 to 
more than 60 million in 1950. A striking development of these years is the 
increase in the number of women regularly employed. In 1950 there were some 
16% million women in the labor force, as compared with 5,000,000 in 1910. 
Almost 27% of all females 10 years old and over were gainfully employed in 
1950; two out of five of these workers were married women. Children, on the 
other hand, were a much smaller proportion of the labor force at the midcentury 
than in the decade after 1900. Before 1910 some 20% of all children between 
the ages of 10 and 15 in the United States were gainfully employed, but by 1930 
the proportion had already fallen to less than 6%. The percentage of older 
workers has also fallen; in 1900 almost 40% of all persons 65 years old and 
over were gainfully employed, but in 1950 the proportion was less than 25 %. 1 

^Lee, Miller, Brainerd, and Easterlin, Population Redistribution and Economic Growth, 
United States, 1870-1950 (Philadelphia: American Philosophical Society, 1957), pp. 598-99. 
In 1950 slightly higher proportions of females 10 years old and over were employed in the 
textile states of New England and the Southeast than in the rest of the country; Rhode 
Island with 33% was the highest; the lowest was West Virginia with only 17.7%. Oppor- 
tunities for female employment created by the offices of the Federal Government can be 
seen in the District of Columbia, where the proportion reached as high as 46%. 

/ 233 



234 / Regional Economic Development: 1870-1950 

Table 88. U. S. labor force, distribution by major industry group, 1910, 
1930, and 1950 



Year 


Labor force 


% 
Resource 


% 
Manufactures 


Services 


1910 
1930 
1950 


38,167,336 
48,829,920 
60,200,847 


35.62 
23.97 
13.59 


27.92 
28.90 
23.92 


36.46 
47.13 
62.49 



Source: Appendix Tables A1-A7. 



Economic opportunities for Negroes after 1910 accelerated migration out of 
both the rural and the urban Southeast, mainly into the cities of the Middle 
Atlantic and Great Lakes regions. Over the decade 1910-1920 Negro migration 
to the Northeast and West showed a net gain for these regions of 334,000. 2 At 
the beginning of the century more than 86% of all gainfully occupied Negroes 
were employed in agriculture or personal and domestic service. By 1920, that 
proportion had fallen to 67%, and almost one-third of all working Negroes 
were engaged in manufactures, trade, and transportation, though mostly in the 
lower-paid jobs requiring little skill. 3 Even more substantial opportunities for 
Negro workers opened up in the urban-industrial centers of the Northeast and 
Far West in the decade after 1940. 

The growth of opportunities for both women and Negroes during and after 
World War I undoubtedly reflected not only the rise of service industries but 
also the decline in immigration from Europe, which had contributed so much 
to the labor force in prewar years. The Census of 1910 showed that the propor- 
tion of the foreign born in the total population was about 14%, whereas in the 
labor force it was 20%. Immigrants comprised about 45% of the labor force 
in mining, 36% in manufactures, and 25% in the transportation industries in 
1910. About 70% of all garment workers were foreign born, as were over half 
of all workers in the iron and steel mills, oil refineries, slaughtering and meat- 
packing plants, furniture factories, leather tanneries, and textile mills. In the 
important transportation industries half the longshoremen and laborers on the 
electric and steam railroads were foreign born. 4 About half of the gainfully 
employed foreign-born women in 1910 were engaged in personal and domestic 
service and about one-fourth in manufactures. 

2 U. S. Department of Commerce, Bureau of the Census, 14th Census of the United States, 
1920: Population, Vol. II (Washington: Government Printing Office, 1922), Table 9, p. 616. 

3 L. V. Kennedy, The Negro Peasant Turns Cityward (New York: Columbia University 
Press, 1930), pp. 32-33; E. J. Scott, Negro Migration During the War (New York: Oxford 
University Press, 1920), pp. 13-18; L. J. Greene and C. G. Woodson, The Negro Wage 
Earner (Washington, D. C: The Association for the Study of Negro Life and History, Inc., 
1930), pp. 339-44. 

4 W. M. Leiserson, Adjusting Immigrant and Industry (New York: Harper and Brothers, 
1924), pp. 7-14; J. Jenks and W. J. Lauck, The Immigration Problem (New York: Funk 
and Wagnalls Co, 1922), pp. 148-52. 



Regional Adjustments within National Economy, 1910-1950 (2) / 235 

The majority of these immigrants and their children lived in urban-industrial 
centers; 71% of the entire foreign white stock made up almost half — 48% — of 
the nation's urban population. In 1920 immigrants comprised over 25% of the 
total population in New England, 22% in the Middle Atlantic region, and 15% 
in the Great Lakes region. Together with their children (foreign white stock) 
they made up 61%, 54%, and 42%, respectively, of the populations in these 
three regions. In the Far West 20% of the population was foreign born and 
44% of foreign white stock. The comparable figures for the Great Lakes region 
were 15% and 42% and for the Mountain region 14% and 36%. In the 
Plains region only 11% of the population were of foreign birth, but 37% were 
of foreign white stock, and these were more concentrated in rural areas than the 



Table 89. Rank order of states with greatest relative concentration of foreign 
born white and of foreign white stock, 1920 and 1950 

(proportion in the white population of the U. S. — 100) 





Foreign 


born 






Foreign white 


: stock 




1920 




1950 




1920 


1950 




R. I. 


202 


N. Y. 


239 


N. D. 


196 


R. I. 


202 


Mass. 


196 


Mass. 


206 


Minn. 


186 


Conn. 


195 


Conn. 


192 


Conn. 


203 


Wis. 


176 


Mass. 


193 


N. Y. 


189 


R. I. 


194 


R. I. 


174 


N. D. 


181 


Ariz. 


185 


N. J. 


186 


Mass. 


164 


N. Y. 


177 


N. J. 


168 


N. H. 


145 


Conn. 


164 


N. J. 


175 


Nev. 


145 


Mich. 


135 


S. D. 


154 


Minn. 


157 


Calif. 


144 


Calif. 


132 


N. Y. 


153 


N. H. 


144 


N. H. 


143 


111. 


129 


N. J. 


149 


Wis. 


142 


N. D. 


142 


Wash. 


110 


Mich. 


140 


111. 


135 



Source: 17th U. S. Census, 1950, Vol. IV, Special Reports, Pt. 3, Chap. A., Tables 13 and 
12; E. P. Hutchinson, Immigrants and Their Children, 1850-1950 (New York: John 
Wiley and Sons, 1956), p. 29. 



foreign stock in any other part of the country. The Southeast and Southwest 
had been least affected by the flood of European immigration, but proximity to 
the Mexican border had raised the Southwest's proportion of foreign white stock 
to more than 11%. In 1920, in fact, Arizona had the highest relative concen- 
tration of foreign-born residents outside the Middle Atlantic and New England 
states. 5 

Between 1920 and 1950 the proportion of foreign born in the total population 
fell from 14% to 7.5% and the proportion of foreign white stock from 24% to 
14.4%. Table 89 shows the concentration of these two groups relative to total 

5 N. Carpenter, Immigrants and Their Children, 1920, Census Monographs VII (Washing- 
ton: Government Printing Office, 1927), pp. 308-9; E. P. Hutchinson, Immigrants and 
Their Children, 1850-1950 (New York: John Wiley and Sons, 1956), pp. 22-62. 



236 / Regional Economic Development: 1870-1950 

white population in various states during this period. The foreign born are 
chiefly concentrated in the New England and Middle Atlantic states in both 1920 
and 1950 and to a lesser extent in the Far West. 

The situation on the foreign white stock is somewhat different. In 1920 the 
highest relative concentration of immigrants and their native-born children was 
in the Plains states of North Dakota and Minnesota. By 1950 the relative con- 
centration of foreign white stock is understandably highest in New England, 
where most foreign-born parents had concentrated in the twenties. Meanwhile 
the Middle Atlantic states of New York and New Jersey have tended to achieve 
greater concentrations than the Plains states, and the lowest concentrations 
among leading states are in Wisconsin and Illinois. It follows, of course, that 
immigrants and their children are a much less important part of the labor force 
during the decade of World War II than during the decade of World War I. 
Nevertheless, with the exception of the Plains region, immigrants and their 
children remained concentrated in manufacturing and service occupations, 
which proliferate in the more urban-industrial regions of the country. 6 



Regional Distribution of the Labor Force 

The rate of movement of the labor force to the regions west of the Mississippi 
River has been less than half as great in the period since 1910 as before 1910. 
Also, the westward movement of the labor force, like that of population gen- 
erally, has become more highly selective as among the four western regions. 
Table 94 shows that the Mountain and Plains regions declined relatively over 
the period 1910-50 and the Southwest had only a moderate increase, while the 
Far West almost doubled its share of the national labor force. In the Northeast, 



Table 90. Regional distribution of U. S. labor force, 1910, 1930, and 1950 

Region 1910 1930 1950 

United States 



New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 



38,167,337 


48,829,920 


60,200,847 


100% 


100% 


100% 


7.64 


7.03 


6.47 


23.56 


24.52 


23.46 


19.02 


20.70 


20.70 


24.51 


21.01 


20.74 


11.66 


10.35 


9.22 


6.19 


6.84 


7.12 


2.24 


2.14 


2.21 


5.18 


7.41 


10.08 



Source: Appendix Table A-l. 

C E. P. Hutchinson, op. cit., pp. 197-218; Lee, Miller, Brainerd, and Easterlin; op. cit. 
pp. 349-60. 



Regional Adjustments within National Economy, 1910-1950 (2) / 237 

New England continued its gradual relative decline throughout these years. 
But the Middle Atlantic and the Great Lakes regions increased their shares 
until 1930, and the Great Lakes managed to hold its own after 1930. Their 
combined share of the national labor force was greater in 1950 than in 1910. 
Read in conjunction with the relative declines of the Southeast, Mountain, and 
Plains regions since 1910, the remarkable buoyancy of the Great Lakes and 
Middle Atlantic regions is evidence of some northward and eastward movement 
of the working population, though the trend has been relatively more into the 
Great Lakes region than the seaboard region. Thus in the period since 1910 
only the Far West, Southwest, and Great Lakes regions have achieved positive 
rates of increase in both population and labor force (Table 91). But whereas 



Table 91. Percentage change in regional shares of U. S. population and labor 
force. 1910-1950 



Region Population Labor force 

New England - 13.20 -15.31 

Middle Atlantic - 2.91 - 0.42 

Great Lakes + 1.66 + 8.83 

Southeast - 6.31 -15.38 

Plains - 26.24 -20.93 

Southwest + 14.05 +15.02 

Mountain + 5.00 - 1.34 

Far West +109.03 +94.59 

Source: 17th U. S. Census, 1950, see supra, Table 1, note; Appendix Table A-l. 



the Far West increased its share of population more than its share of the labor 
force, the Great Lakes region increased its share of the labor force considerably 
more than its share of total population. The Southwest's increase was about the 
same in both. 



Resource Activities in the Regions, 1910-1950 

The share of resource activities in the total labor force of the United States 
declined by 61% between 1910 and 1950, compared with a decline of 33% in 
the 1870-1910 period. Nevertheless, the country's natural resources still provide 
the basic materials for all other economic activities, and it is only recently that 
dependence on foreign materials for strategic purposes has become a matter of 
public concern. The declining share of resource activities is the logical outcome 
of rising labor productivity in all the major resource sectors and, during the 
present century at least, it also reflects a decreasing use of materials per unit of 
manufactured output. 

Within this secular decline of primary activities countrywide, there have been 



238 / Regional Economic Development: 1870-1950 

significant regional variations, reflecting in part the differential experience of 
the four component resource sectors: agriculture, mining, fisheries, forestry and 
logging (Table 92). Between 1910 and 1950 the combined share of the three 
Northeastern regions in all primary employment fell by 10%, as did the share 
of the Southeast also ; at the same time, the combined share of the four Western 
regions rose by 23%. The Southeast's relative decline over these years, however, 
was only in agriculture. In the other resource sectors it has greatly expanded its 
share of resource employment since 1910, and in the mining and forest sectors, 
in particular, its growth has been much more striking than that of the greater 
West. Within the West, moreover, it is the Far West that has contributed most 
of the expansion; with the significant exception of agriculture, the shares of the 
Plains and Mountain regions have generally declined. 



Table 92. Regional distribution of all U. S. resource industries labor force, 
1910, 1930, and 1950 



Region 



1910 



1930 



1950 



United States 



New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 



13,596,815 


11,706,790 


8,184,509 


100% 


100% 


100% 


2.29 


2.06 


1.86 


9.68 


8.73 


8.12 


15.05 


13.85 


14.26 


42.34 


39.95 


38.06 


14.07 


15.05 


16.87 


10.41 


11.55 


10.39 


2.45 


3.32 


3.53 


3.71 


5.49 


6.91 



Source: Appendix Table A2-A5. 



Table 93. Regional distribution of each resource industry's labor force, 
1910-1950 



Agriculture 



Region 


1910 


1920 


1930 


1940 


1950 


United States 


12,389,840 


10,665,812 


10,471,998 


8,700,376 


6,962,779 




100% 


100% 


100% 


100% 


100% 


New England 


2.27 


2.10 


2.03 


1.76 


1.76 


Middle Atlantic 


7.54 


7.03 


6.54 


6.27 


6.27 


Great Lakes 


14.74 


14.95 


13.86 


14.49 


14.95 


Southeast 


44.50 


41.93 


41.36 


40.80 


38.28 


Plains 


14.67 


15.64 


16.20 


17.10 


19.00 


Southwest 


11.01 


11.17 


11.89 


11.36 


9.86 


Mountain 


2.06 


2.97 


3.09 


2.98 


3.39 


Far West 


3.21 


4.21 


5.03 


5.24 


6.49 



Regional Adjustments within National Economy, 1910-1950 (2) / 239 



Mineral Extraction 



Region 


1910 


1920 


1930 


1940 


1950 


United States 


965,169 


1,090,223 


984,323 


913,000 


929,421 




100% 


100% 


100% 


100% 


100% 


New England 


.94 


.45 


.67 


.51 


.52 


Middle Atlantic 


36.68 


32.17 


32.22 


26.47 


22.46 


Great Lakes 


19.61 


19.84 


14.71 


12.43 


11.61 


Southeast 


16.73 


22.53 


26.22 


30.26 


34.87 


Plains 


8.40 


6.80 


5.40 


5.08 


5.16 


Southwest 


4.33 


8.49 


9.86 


12.89 


16.24 


Mountain 


7.49 


5.88 


5.57 


5.75 


4.92 


Far West 


5.82 


3.84 


5.35 


6.61 


4.22 


Forestry and Logging 


Region 


1910 


1920 


1930 


1940 


1950 


United States 


173,531 


217,378 


177,189 


186,688 


215,432 




100% 


100% 


100% 


100% 


100% 


New England 


5.23 


10.29 


6.08 


7.53 


6.17 


Middle Atlantic 


7.77 


7.16 


5.36 


4.43 


4.70 


Great Lakes 


13.00 


16.10 


11.05 


6.88 


6.12 


Southeast 


35.13 


27.90 


32.73 


43.91 


44.15 


Plains 


7.32 


9.27 


5.25 


3.83 


3.61 


Southwest 


4.93 


3.63 


5.09 


5.03 


4.38 


Mountain 


3.31 


4.74 


5.76 


3.38 


3.11 


Far West 


23.31 


20.91 


28.68 


25.01 


27.76 


Fisheries 


Region 


1910 


1920 


1930 


1940 


1950 


United States 


68,275 


52,836 


73,280 


60,027 


76,877 




100% 


100% 


100% 


100% 


100% 


New England 


18.23 


17.15 


15.35 


14.87 


15.80 


Middle Atlantic 


21.21 


17.02 


14.27 


13.57 


12.36 


Great Lakes 


11.03 


8.33 


8.34 


7.33 


6.06 


Southeast 


31.71 


33.93 


39.67 


41.55 


39.75 


Plains 


2.30 


2.42 


2.55 


1.66 


1.93 


Southwest 


1.71 


1.27 


2.12 


3.38 


4.37 


Mountain 


.22 


.17 


.24 


.50 


.68 


Far West 


13.59 


19.71 


17.46 


17.14 


19.05 



Source: Appendix Tables A2-A5. 



In the agriculture sector, both the Northeast and the Southeast declined rela- 
tive to the West after 1910 (Table 93). But the Southeast's share fell off by 
about 11%, and that of the Northeast by only 6%. In terms of labor force, at 
least, the agricultural strength of the Northeast lay in the Great Lakes rather 



240 / Regional Economic Development: 1870-1950 

than in the other two regions. The reverse side of this relative agricultural 
decline in the East is, of course, the rising share of agricultural activities in the 
greater West: the Plains, Far West, and Mountain regions each contributed to 
the over-all increase in employment shares of about one-quarter between 1910 
and 1950. 

In the minerals sector the Northeast's decline is very marked, its relative share 
falling off by almost 40% over the four decades. The decline in coal mining 
affected the Middle Atlantic region somewhat more than the Great Lakes region, 
but both had substantially smaller shares in 1950 than in 1910. The West 
increased its share of mining employment by about one-sixth over these years; 
most of this increment accrues to the Southwest, since other Western regions 
had smaller shares of the nation's mining activity in 1950 than in 1910. In 
contrast, the share of the Southeast in the nation's mining sector more than 
doubled, with much of the increase taking place in the late thirties and forties. 

The Northeast's share of forest and logging employment declined relatively 
by more than one-third between 1910 and 1950. The West's share remained 
relatively unchanged, the decline of forest activities in the Plains states being 
just about offset by the rise in the Pacific Northwest. Almost all of the relative 
gain in forest activities in this period, therefore, was achieved in the Southeast 
which increased its share of forest employment by almost one-fourth. In 1950 
more than 44% of all forest and logging activity in the United States was 
located in the Southeast. 

In the nation's fisheries, also, the Northeast suffered relative decline in the 
period from 1910 to 1950, its employment falling off relative to the rest of the 
nation by almost one-third. New England, however, made a larger relative con- 
tribution to national employment in the fisheries than in any other resource 
sector. The Southeast raised its share of fishing employment during these years 
by about one-fourth, and the West its share by almost 47%. Most of the West's 
gain came in the coastal states of the Far West, but a not insignificant share 
was contributed by fishermen of the Southwest sailing from ports on the Gulf 
Coast of Texas. 

The application of "shift analysis" to resource employment in the states 
enables us to determine more precisely which parts of the country contributed to 
growth in the various resource sectors and conversely which suffered greatest 
relative declines. 



AGRICULTURE 

After 1910 the most notable trend in agricultural employment is absolute 
decline: from 10,471,998 in 1910 to 6,962,779 in 1950. Though several states 
achieved absolute increases in farm employment during particular decades, only 
five states had absolute net increases over the 40-year period as a whole (Table 
94). Four of these have actually experienced absolute decline since 1930, 
although their 1950 total still represented a net increase over 1910; one, Mon- 
tana, has been in absolute decline since 1920. The other five among the ten 
states with the greatest net upward shifts all have had absolute declines since 



Regional Adjustments within National Economy, 1910-1950 (2) / 241 




Absolute upward shift 
y'/A Relative upward shift 
I I Absolute downward shift 

Figure 35. Net Shift in Agricultural Labor Force, 1910-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 15.2. Total net shift as % of 1910 agricultural labor 
force: 6.7. Source: Appendix Table D. 



Table 94. 
1950 



Principal net shifts in agricultural labor force among states, 1910- 





Percentage net 




Percentage 


net 






upward shift 




downward s 


hift 




California 




21.40 


Georgia 




15.74 


Florida 




6.69 


Alabama 




13.17 


Arizona 




3.11 


S. Carolina 




9.73 


Montana 




3.04 


Texas 




8.36 


Idaho 




3.01 


Mississippi 




7.39 


Minnesota 




12.60 


New York 




5.15 


Wisconsin 




1028 


Arkansas 




4.77 


Iowa 




10.19 


Pennsylvania 




4.71 


Nebraska 




4.49 


Oklahoma 




4.69 


N. Carolina 




3.88 


Louisiana 




4.12 



States in italics experienced relative net increases only. 
Source: Appendix Table A-2. 



1910; but because their rate of decline in each case was slower than for the 
nation as a whole, they appear in 1950 with relative net increases for the 40-year 
interval. Twenty-six states experienced absolute declines in farm employment, 
and all but five of the remaining twenty-two had only relative increases. 



242 / Regional Economic Development: 1870—1950 

In the East, Florida was the only state with absolute increase in farm employ- 
ment during the years 1910-50, but three others — Wisconsin, North Carolina, 
and New Jersey — ended the period with relative increases. 7 All New England 
states showed absolute declines, although their combined declines accounted 
for only 4.3% of the total for the nation. 

The heaviest downward shift occurred in the Southeast, especially in South 
Carolina and westward to Arkansas and Louisiana, reflecting the decline of 
cotton and other labor-intensive types of row-cropping as well as the effects of 
farm mechanization. Much of this net downward shift in the Southeast, inci- 
dentally, took place after 1930, although some of it occurred during World 
War I. Texas and Oklahoma in the Southwest together accounted for more than 
13% of the nation's total net downward shift; the causes here were much the 
same as in the Southeast. However, owing to the expansion of truck farming 
and certain field crops, Arizona was third only to California and Florida among 
states registering absolute net increases over the period. Montana and Idaho in 
the Mountain region, the other two states with absolute net upward shifts, seem 
to owe their rising level of farm activities to the westward encroachment of the 
wheat belt during World War I. 



MINING 

Unlike agriculture, mineral production taken as a whole, and for the whole 
period 1910-1950, has been a more or less stabilized sector of resource employ- 
ment (see Table 93). There have been, however, a number of ups and downs 
in employment. There was some falling-off in mining activities between the two 



Table 95. Principal net shifts in mining labor force among states, 1910-1950 





Percentage 


net 






Percentage 


net 






upward shift 






downward shift 




Texas 






28.01 


Pennsylvania 




42.16 


West Virginia 






24.66 


Ohio 






7.51 


Kentucky 






16.87 


Illinois 






7.40 


Oklahoma 






9.78 


Michigan 






6.75 


Louisiana 






7.83 


Colorado 






5.83 


Virginia 






5.55 


Missouri 






5.58 


New Mexico 






1.58 


Iowa 






4.05 


Mississippi 






1.21 


Montana 






3.21 


Utah 






.83 


Indiana 






2.78 


Arkansas 






.70 


Washington 




2.42 



Source: Appendix Table A-3 



7 Relative increases of the agricultural labor force in North Carolina and New Jersey 
are probably accounted for by increases in tobacco and vegetable production, respectively. 
Wisconsin's relative increase reflects the stability of labor requirements in dairying. 



Regional Adjustments within National Economy, 1910—1950 (2) / 243 




Figure 36. Net Shift in Mining Labor Force, 1910-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 818.8. Total net shift as % of 1910 mining labor force: 
30.3. Source: Appendix Table D. 



World Wars and the trend was not reversed until after 1940. Between 1940 and 
1950 mineral employment rose by a little more than 1%. 8 

The period 1910-1950 as a whole is marked by the rise of mineral fuels t« 
dominance in the nation's mining activity. By 1950 about three-quarters of all 
mining employment originated in either coal or petroleum; metallic and other 
non-metallic mining had decreased correspondingly. Developments in a few 
Southeastern and Southwestern states, moreover, account for nearly 93% of the 
total net upward shift in all mining employment since 1910 (Table 95). Minor 
upward shifts, all accounting for less than 1% of the total, also occurred in 
three western Plains states (mostly in Kansas) and in two Mountain states 
(Wyoming and Utah). Absolute net downward shifts were registered by all the 
Northeastern states (more than 42% of the nation's total in Pennsylvania alone) 
and by all the Far Western states except California, where the upsurge of 
petroleum activities between 1920 and 1940 offset the sharp decline in metal 
mining. The two great mining states of the Mountain region, Montana and 
Colorado, also experienced sizable absolute net downward shifts in mining labor 
force, amounting to more than 9% of the national total over the 40-year period. 



8 A more recent figure from the Census of Manufactures 1954, however, suggests that 
the number of production workers, at least, is again in decline — nearly 9 per cent since 
the 1939 Census of Manufactures. 



244 / Regional Economic Development: 1870-1950 

For practical purposes, therefore, the growth of petroleum, natural gas, and 
coal in Texas, Louisiana, West Virginia, Kentucky, and Virginia is the sum of 
the growth in mining employment achieved in recent decades, though World 
War II brought a revival to the Minnesota iron fields after three successive 
decades of steady decline. On the other hand, the decline of coal mining in the 
Middle Atlantic and Great Lakes states is the most important factor in the 
downward shift. 



FORESTRY AND LOGGING 

The record of employment in forestry and logging in the 1910-1950 period 
resembles that of mining, but recovery in this sector between 1940 and 1950 was 
much more substantial, amounting to more than 15%. In the past four decades, 
logging activities have been increasingly concentrated in the Far West (almost 
exclusively in Oregon) and the Southeast (Table 96). Eight of the ten principal 
growth states are located in the Southeast; these states contributed almost 55% 
of the total upward shift in logging activities between 1910 and 1950. 

World War I brought a revival of logging activities in some of the northern 
New England states, New York, and states bordering Lake Superior, but there- 
after the camps moved off to the Pacific Northwest, the Southeast, and the 




Absolute upword shift 
fcv/--| Relative downward shift 
I I Absolute downward shift 



Figure 37. Net Shift in Forest Labor Force, 1910-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 132.0. Total net shift as % of 1910 forest labor force: 
31.9. Source: Appendix Table D. 



Regional Adjustments within National Economy, 1910-1950 (2) / 245 

Southwest. During the depression decade more than 68% of the entire net 
upward shift in logging activity took place in Georgia and Florida, but by 
World War II employment shifted back again to Oregon and California in the 
Far West and to the Carolinas and Mississippi/ 9 During this century the forest 
industries have been profoundly affected, on the demand side, by the greatly 
increased domestic pulpwood requirements and, on the supply side, by policies 



Table 96. Principal net shifts in forest labor force among states, 1910-1950 





Percentage 


net 




Percentage 


: net 






upward si 


lift 




downward 


shift 




Oregon 






35.02 


Washington 




17.87 


Georgia 






22.21 


Michigan 




17.12 


S. Carolina 






9.37 


Minnesota 




10.74 


N. Carolina 






5.04 


Wisconsin 




10.67 


Maine 






4.59 


Pennsylvania 




10.44 


Mississippi 






4.10 


West Virginia 




9.23 


Florida 






3.96 


Louisiana 




8.15 


Alabama 






3.89 


Missouri 




4.26 


Virginia 






3.50 


Tennessee 




2.63 


Arkansas 






2.33 


Oklahoma 




1.91 



Source: Appendix Table A-4. 



of conservation and reforestation. Much of the enduring strength of the South- 
east and the northern New England states can be attributed to conservation 
policies supported by government and industry. Unfortunately, recognition of 
the economic advantages of sustained yield practices came too late to arrest the 
devastation of the great forests of Michigan, Wisconsin, and Minnesota. These 
three states, as a consequence, accounted for more than 38% of the total net 
downward shift in logging activities between 1910 and 1950; most of their 
decline occurred in the years after 1920. 



9 There has been, in fact, a curious cycle in logging activity over the years since 1910 : 
many of the states which experience the largest net upward shifts during one decade often 
experience the largest net downward shifts during the ensuing decade and vice versa. 
Thus the largest net upward shifts 1910-20 occurred in Maine, Minnesota, Wisconsin, 
and New York; the largest net downward shifts from Washington, Pennsylvania, West 
Virginia and Louisiana. Over the decade 1920-30 the largest net upward shifts were 
registered in Oregon, Washington, Florida, and Mississippi; the largest downward shifts 
from Minnesota, Wisconsin, Maine, and Michigan. During 1930-40 the largest upward 
shifts took place in Georgia, Florida, Alabama, and South Carolina; the largest downward 
shifts from Washington, Michigan, Idaho, and Minnesota. 1940-50 saw the largest net 
upward shifts in Oregon, California, North and South Carolina, the largest net downward 
shifts from Florida, Washington, Georgia, and New Hampshire. 



246 / Regional Economic Development: 1870-1950 



FISHERIES 

Most of the trends evident in fishing employment during the late nineteenth 
century have intensified since 1910. The heaviest net downward shifts during 
the 1910-1950 period have been in the New England and Middle Atlantic regions 
and Chesapeake Bay areas, heavy net upward shifts into Far Western and in cer- 
tain Gulf states (Table 97). Since the 1930's, downward shifts in the Great 
Lakes fisheries, especially in Illinois, have also contributed to the relative 




Figure 38. Net Shift in Fishing Labor Force, 1910-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 194.0. Total net shift as % of 1910 fishing labor force: 
24.4. Source: Appendix Table D. 



Table 97. Principal net shifts in fishing labor force among states, 1910-1950 





Percentage net 




Percentage 


net 






upward shift 




downward shift 




Louisiana 




23.40 


Maryland 




26.12 


California 




22.84 


Virginia 




14.85 


Florida 




20.94 


Illinois 




14.31 


Texas 




11.41 


New Jersey 




10.18 


Washington 




5.80 


Massachusetts 




6.33 


Mississippi 




5.10 


Michigan 




3.82 



Source: Appendix Table A-5. 



Regional Adjustments within National Economy, 1910-1950 (2) / 247 

decline in the Northeast. The greatest gains were achieved during this period by 
Louisiana, California, Florida, and Texas; together they accounted for more 
than 78% of the total net upward shift over these years. Far Western and 
Gulf fisheries have been providing about half the total value of the fishing catch 
in recent years, which indicates that canned and frozen seafoods, together with 
menhaden and other meal-oil species, have become a much more lucrative sector 
of the industry than the market fish brought into ports along the North Atlantic 
and Great Lakes shores. The revival of the Maine fisheries in New England after 
1940 was also probably connected with the development of new processing 
facilities in that state. 



The Value of Extracted Resources 

In terms of the gross value of resources extracted in the various regions, 
the most striking change has been the almost complete reversal in the positions 
of the Northeast and greater West between 1870 and 1950. In 1870 the three 
Northeastern regions contributed 58.2% to the gross value of the nation's 
extracted resources. By 1950 that proportion had fallen to 25.5%. The share of 
the four Western regions rose during these years from 16.6% in 1870 to 54.4% 
in 1950. The relative share of the Southeast has remained almost unchanged 
since the 1880's. 

Since 1910, however, the rate of redistribution between the Northeast and the 
West has slowed markedly. Between 1870 and 1910 the Northeast's share in 
total resources value declined by more than 39% ; but since 1910 the decline has 
been less than 28%. And the share of the West rose by more than 153% 
between 1870 and 1910, but by only 29% after 1910. 

In the earlier period the Southeast's share declined by more than 11%, but 
after 1910 the decline slowed to 10%. 



Table 98. Regional distribution of value of primary resources extracted, 1910, 
1930, and 1950 



Region 


1910 


1930 


1950 


United States 


$10,512,395* 


$17,426,866* 


$35,030,981* 




100% 


100% 


100% 


New England 


3.2 


2.6 


2.1 


Middle Atlantic 


11.6 


11.8 


8.5 


Great Lakes 


20.6 


15.0 


14.9 


Southeast 


22.4 


21.0 


20.1 


Plains 


24.5 


18.7 


18.6 


Southwest 


7.6 


13.9 


16.9 


Mountain 


4.1 


5.6 


5.6 


Far West 


6.0 


11.4 


13.3 



*000's of current dollars. 
Source: Appendix Table B-l. 



248 / Regional Economic Development: 1870-1950 

Table 98 reveals that New England's decline in resources value continued 
throughout the 1910-1950 period. The Great Lakes region, on the other hand, 
declined by one-fourth between 1910 and 1930, but thereafter retained an 
almost constant share of the greatly increased total value of extracted resources. 
In the years down to 1930 the Middle Atlantic region contributed stability to 
the share of the Northeast, but thereafter it too went into relative decline. 
Among the Western regions, the Far West and Southwest contributed most to 
the growth of extracted resources value; between 1910 and 1950 both these 
regions more than doubled their shares, while the Mountain region leveled off 
after 1930 and the Plains region continued its decline from the peak reached in 



Table 99. Percentage distribution of value of all resources extracted, by major 
resource industry and great region, 1910, 1930, and 1950 



Resource industry 



1910 



1950 



All resource indust 


ries 


$14,827,073* 


$20,667,245* 








100% 




100% 




Agriculture 




80.80 




62.95 




Mineral extraction 




11.78 




30.20 




Forestry & logging 




6.90 




5.96 




Fisheries 




.52 




.89 




Great region 


: 


Northeast 


Southeast 


West 


United States 


Agriculture 












1910 


27.19 


17.64 


35.97 




80.80 


1930 


17.64 


13.54 


32.01 




63.19 


1950 


17.74 


11.09 


34.12 




62.95 


Mineral extraction 












1910 


6.01 


1.52 


4.25 




11.78 


1930 


10.44 


4.46 


14.74 




29.64 


1950 


6.70 


6.83 


16.67 




30.20 


Forestry and logging 










1910 


1.'93 


3.14 


1.83 




6.90 


1930 


1.02 


2.82 


2.69 




6.53 


1950 


.71 


2.00 


3.25 




5.96 


Fisheries 












1910 


.31 


.13 


.08 




.52 


1930 


.32 


.16 


.17 




.65 


1950 


.31 


.23 


.35 




.89 


*000's of 1929 dollars. 










Source : Appendix 


Tables B1-B5. 











Regional Adjustments within National Economy, 1910-1950 (2) / 249 

Table 100. Regional distribution of value of resources extracted, by major 
resource industry, 1910, 1930, and 1950 

Agriculture 



Region 


1910 


1930 


1950 


United States 


$8,494,231* 


$11,011,329* 


$22,052,255* 




100% 


100% 


100% 


New England 


2.89 


2.86 


2.41 


Middle Atlantic 


9.13 


8.26 


7.44 


Great Lakes 


21.63 


16.81 


18.34 


Southeast 


21.83 


21.42 


17.62 


Plains 


28.33 


25.83 


25.80 


Southwest 


8.04 


10.59 


11.71 


Mountain 


3.17 


5.35 


5.62 


Far West 


4.98 


8.88 


11.06 


Mineral Extraction 


Region 


1910 


1930 


1950 


United States 


$1,238,415* 


$5,164,968* 


$10,579,973* 




100% 


100% 


100% 


New England 


1.40 


.95 


.41 


Middle Atlantic 


30.45 


21.18 


11.66 


Great Lakes 


19.18 


13.09 


10.11 


Southeast 


12.90 


15.03 


22.62 


Plains 


10.52 


7.61 


7.46 


Southwest 


6.15 


23.38 


30.67 


Mountain 


11.46 


6.73 


5.96 


Far West 


7.94 


12.03 


11.11 



: 000's of current dollars. 

Forestry and Logging 



Region 1910 1930 1950 



United States 



New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 



$725,737* 


$1,137,180* 


$2,086,249* 


100% 


100% 


100% 


7.78 


4.48 


4.24 


7.36 


3.55 


3.47 


12.88 


7.56 


4.14 


45.48 


43.18 


33.56 


5.18 


2.12 


2.13 


4.95 


4.74 


3.86 


2.23 


3.25 


3.98 


14.14 


31.12 


44.62 



250 / Regional Economic Development: 1870-1950 

Table 100. Regional distribution of value of resources extracted, by major 
resource industry, 1910, 1930, and 1950 (continued) 

Fisheries 



Region 


1910 


1930 


1950 


United States 


$54,012* 


$113,389* 


$312,504* 




100% 


100% 


100% 


New England 


28.03 


25.64 


19.38 


Middle Atlantic 


22.26 


16.63 


12.58 


Great Lakes 


9.33 


7.36 


3.33 


Southeast 


25.57 


24.04 


25.66 


Plains 


1.32 


3.32 


.82 


Southwest 


.83 


.93 


3.72 


Mountain 


— 


— 


— 


Far West 


12.66 


22.08 


34.51 



*000's of current dollars. 
Source: Appendix Tables B1-B5. 



Table 101. Distribution of labor force in resource industries and value of 
extracted resources, by great region, 1910, 1930, and 1950 







1910 


1930 






1950 


Region 


Labor 
force 


Value 


Labor 
force 


Value 


Labor 
force 


Value 


United States 


100% 


100% 


100% 


100% 


100% 


100% 


Northeast 
Southeast 
West 


27.0 
42.3 
30.7 


35.4 
22.4 
42.2 


24.6 
40.0 
35.4 


29.4 
21.0 
49.6 


24.2 
38.1 

37.7 


25.5 
20.1 

54.4 



Source: Appendix Tables A-l, B-l. 



1910. It is evident that the relative eclipse of agriculture and the rising output 
of liquid fuels (oil and natural gas) underlies this continuing regional trans- 
formation (Tables 99 and 100) . 

The comparison of regional shares of total resource value with shares in 
primary employment shown in Table 101 throws further light on changes in the 
resource sector since 1910. Thus the Southeast's limited decline in its resource 
employment share can be seen to have been accompanied by a proportionate 
decline in its share of the gross value of extracted resources. The greater North- 
east suffered a much greater relative loss in resource value, with only a small 
decline in labor force. On the other hand, the moderately rising share of the 



Regional Adjustments within National Economy, 1910-1950 (2) / 251 

West in all primary employment was accompanied, as in the period before 1910, 
by a much more than proportionate increase in its share of extracted resources 
value. In recent decades, however, it is the Far West and Southwest rather than 
the Plains or the Mountain regions which have made the greatest contributions 
to the value of resources extracted in the greater West. In short, labor produc- 
tivity in the four primary resource sectors has continued to diverge, and within 
each sector some activities result in high values per unit of labor input, others 
in low. Recently, the Southwest and Far West have been developing resources 
which require relatively low labor inputs for high-value outputs, while the 
reverse has been true of the Southeast. Moreover, during the present century at 
least, the Plains and Mountain regions have been moving more toward the pat- 
tern of the Southeast and have been unable to develop the more lucrative types 
of resource activity characteristic of the Southwest and Far West. 



Manufacturing Activities in the Regions 1910-1950 

As in the earlier period, the most striking feature of American manufacturing 
developments after 1910 was the enduring strength of the urban-industrial 
Northeast. In spite of the decline in the Northeast's natural riches, the combined 
share of the Middle Atlantic, New England, and Great Lakes regions in manu- 
facturing employment declined by only 2.8% between 1910 and 1950. In the 
forty years before 1910, their combined share had declined by more than 13%. 
It has been the Great Lakes region rather than the New England and Middle 
Atlantic regions, however, which has given the manufacturing Belt its strength; 
between 1910 and 1950 the share of the Great Lakes region in all manufacturing 
employment rose by almost 28%. 

Outside the Northeast the most notable relative gains have been registered in 
the Far West and Southeast during these years (Table 102). The Southwest 
gained until 1930 but had dropped again by 1950. In the Plains and Mountain 
regions, manufacturing has, with certain exceptions, been in relative decline. 
Taken as a whole, the greater West, like the Northeast, has a slightly smaller 
share of country-wide employment in manufactures than it had in 1910. 

Thus, after more than eighty years of continuing industrial transformation, 
more than two-thirds of the nation's manufacturing employment is still located 
in the Northeastern industrial belt stretching from southern New England south 
to the Potomac and west to the Mississippi. Manufacturing, therefore, remains 
in 1950 much more localized than either resource or service activities. 

Another instructive measure of manufacturing activity is "value-added by 
manufacture." Unlike the category "value of extracted resources," value-added 
by manufacture is a net figure which enables us to determine precisely how 
much of the final value of a product is attributable to the manufacturing process 
as distinct from materials, containers, power, etc. Table 103 presents the 
regional shares of all value-added by manufacture in the nation as a whole. In 
general, changes in each region's share of total value-added by manufacture 



252 / Regional Economic Development: 1870-1950 



Table 102. Regional distribution of manufacturing labor force, 1910, 1930, 
and 1950 



Region 


1910 


1930 


1950 


United States 


10,656,545 


14,110,652 


14,398,854 




100% 


100% 


100% 


New England 


13.42 


10.48 


9.59 


Middle Atlantic 


33.55 


30.31 


29.11 


Great Lakes 


22.61 


25.59 


28.94 


Southeast 


12.69 


14.38 


15.43 


Plains 


8.35 


7.11 


5.68 


Southwest 


2.74 


4.13 


3.28 


Mountain 


1.64 


1.34 


.87 


Far West 


5.00 


6.66 


7.10 



Source: Appendix Table A-6. 



Table 103. Regional distribution of value added by manufacture, 1910, 1930, 
and 1950 



Region 


1910 


1930 


1950 (47) 


United States 


$8,188,527* 


$30,693,709* 


$74,353,602* 




100% 


100% 


100% 


New England 


14.32 


10.35 


9.16 


Middle Atlantic 


36.90 


33.75 


29.87 


Great Lakes 


25.59 


31.63 


31.57 


Southeast 


10.04 


9.86 


12.49 


Plains 


6.42 


5.74 


5.54 


Southwest 


1.49 


2.00 


3.00 


Mountain 


1.21 


.95 


.88 


Far West 


4.03 


5.72 


7.49 



* 000's of current dollars. 
Source : Same as Table 46. 



parallel changes in its share of total manufacturing employment and, as in the 
earlier period, the parallel is closer than that between changes in the shares of 
extracted resources value and in the shares of primary employment. By the late 
1940's, however, two regions still deviate somewhat from the close parallel 
existing in the other six : the Southeast has a considerably larger share of manu- 
facturing employment than of value-added, while the Great Lakes region has a 
larger share of value-added than of manufacturing employment. It is significant 
perhaps that before 1930 the Middle Atlantic region also had a larger share of 



Regional Adjustments within National Economy, 1910-1950 (2) / 253 




Manufacturing Labor Force 



1950 




14,110,652 Employed 



14,398,854 Employed 



S.W. 2,0%- X 



F.W. 5.7 



Value Added by Manufacture 



s.w 





$30,693,709,000 



$74,353,602,000 



Figure 39. Regional Distribution of U. S. Manufacturing Labor Force and 
Value Added by Manufacture, 1930 and 1950. 



Table 104. Distribution of manufacturing labor force and value added by 
manufacture, by great region, 1910, 1930, and 1950 





1910 




1930 




1950 


Region 


Labor 
force 


Value 
added 


Labor 
force 


Value 
added 


Labor 
force 


Value 
added 


United States 


100% 


100% 


100% 


100% 


100% 


100% 


Northeast 
Southeast 
West 


69.58 
12.69 
17.73 


76.81 
10.04 
13.15 


66.38 
14.38 
19.24 


75.73 

9.86 

14.41 


67.64 
15.43 
16.93 


70.60 
12.49 
16.91 



Source: Appendix Table A-6; Table 103. 



254 / Regional Economic Development: 1870-1950 




j ■'■;'; I Relative downward shift 
| | Absolute downward shift 



Figure 40. Net Shift in Manufacturing Labor Force, 1910-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 52.7. Total net shift as % of 1910 manufacturing labor 
force: 18.5. Source: Appendix Table D. 



value-added than of manufacturing employment. The implication of these devel- 
opments, therefore, is that the Great Lakes region is now the nation's most pro- 
ductive manufacturing region (in terms of value-added per worker) and that the 
Southeast is relatively the least productive among those in which manufactures 
are still concentrating. The stress on the relative character of these comparisons 
is particularly important because of the marked increase in labor productivity 
in manufacturing throughout the entire country. 

A comparison of the ratio of manufacturing employment to value-added in 
the three great regions suggests that the disparity between the two is somewhat 
less general today than in 1910 (Table 104). The tendency towards convergence 
is particularly striking if we examine the trend of the two categories in each of 
the regions over the years of depression and war since 1930. Since 1930 only 
the Southeast and Far West have had their shares of value-added rising at a 
faster rate than their shares of manufacturing employment and, what amounts 
to the same thing, only the Plains and Mountain regions have had their shares of 
employment falling at a faster rate than their shares of value-added by manu- 
facture. Broadly speaking, each of these four regions is somewhat more pro- 
ductive relatively (value-added per worker in manufactures) in 1950 than at the 
onset of depression in 1930. On the other hand, the New England and Middle 
Atlantic regions' shares of the nation's value-added has fallen at a faster rate 
than their corresponding shares of manufacturing employment. In the Great 
Lakes region the share of value-added has remained almost stationary since 
1930, but the share of the manufacturing labor force has increased. Broadly 
speaking, therefore, these last three regions are proportionately somewhat less 



Regional Adjustments within National Economy, 1910-1950 (2) / 255 

Table 105. Principal net shifts in manufacturing labor force among states, 
1910-1950 





Percentage 


net 




Percentage 


net 






upward shift 




downward 


shift 




Michigan 






24.53 


New York 




19.24 


California 






18.15 


Massachusetts 




18.45 


N. Carolina 






9.36 


Pennsylvania 




15.17 


Ohio 






8.61 


Missouri 




4.09 


Texas 






5.96 


Minnesota 




3.76 


Illinois 






5.86 


Kansas 




3.38 


Indiana 






5.45 


Iowa 




3.14 


Georgia 






4.50 


Rhode Island 




2.93 


S. Carolina 






3.95 


Maine 




2.87 


New Jersey- 






3.76 


Nebraska 




2.77 



The figures in italics are relative downward shifts only. 
Source: Appendix Table A-6. 






productive in 1950 than in 1930. 10 From an historical standpoint, however, it is 
too early to regard these tendencies as marking any genuine turning point in the 
regional distribution of American manufactures. In the perspective of the entire 
period since 1910 such minor convergence among the regions does not appear 
to have produced any noteworthy redistribution either of manufacturing employ- 
ment or value-added. 

When the developments within particular states in the period 1910-1950 are 
examined, it is evident that there is a tendency for the older urban-industrial 
areas to sustain the greatest relative losses in manufacturing employment (Table 
105 and Figure 40). Nevertheless, it is no longer true that the more rural-agri- 
cultural states necessarily experience the greatest net upward shifts. Except for 
several Southeastern states, most of the relative net increases occurred in such 
states as Michigan, California, Ohio, Indiana, and New Jersey, which were 
already substantial urban-industrial areas before World War I. As a matter of 
fact, apart from California, Texas, and the three Southeastern states — the Caro- 
linas and Georgia — the balance of upward shift was almost wholly confined to 
the Great Lakes region, every one of whose member states registered a net 
upward shift. By 1950 the Great Lakes region was almost level with the Middle 
Atlantic region in its share of the nation's manufacturing employment. 

10 Average value-added per production worker (or per worker employed in manufactures) 
varies from region to region with: 1) the proportion of the manufacturing labor force 
in high, medium, or low value-added industries, and 2) the variation from region to 
region in value-added per worker in the same industry group. A recent Department of 
Commerce study indicates that for a combination of these reasons the Southeast and 
New England rank below the countrywide average value-added per production worker in 
1939 and 1947. The Far West ranks highest above the national average. U. S. Dept. 
of Commerce, Regional Trends in the United States Economy, (Washington: Government 
Printing Office, 1951), pp. 83-87. 



256 / Regional Economic Development: 1870-1950 

Table 106. Principal net shifts in manufacturing labor force among states, 
by decade, 1910-1950 



1910-20 


1920-30 


1930-40 


1940-50 








Percentage net upward shift 








Mich. 


32.38 


Calif. 


22.93 


Mich. 


17.36 


Calif. 


32.54 


Ohio 


20.23 


Mich. 


14.85 


N. C. 


15.92 


Texas 


14.42 


Calif. 


13.76 


Texas 


12.73 


Conn. 


10.17 


Ind. 


8.85 


Texas 


8.22 


111. 


10.05 


N. J. 


9.42 


Wis. 


7.54 


N. J. 


6.52 


N. C. 


8.24 


111. 


8.42 


Ohio 


6.79 


Ala. 


3.72 


Tenn. 


5.04 


Ohio 


7.60 


Minn. 


4.83 


N. C. 


3.64 


Ga. 


4.67 


S. C. 


5.85 


Iowa 


2.75 


Okla. 


2.75 


Fla. 


4.38 


Ga. 


4.21 


Kan. 


2.75 


Fla. 


2.14 


S. C. 


3.81 


Va. 


3.96 


Mich. 


2.48 


Ga. 


1.86 


Okla. 


3.47 


Ind. 


3.76 


Ore. 


2.39 


Share of 
















ten leading 
















states in total 














net upward 
















shift 


95.22 




90.17 




86.67 




85.34 






Percentage net 


downward shift 








N. Y. 


27.82 


Mass. 


28.79 


Texas 


12.08 


N. Y. 


17.38 


Pa. 


13.91 


Pa. 


21.72 


Calif. 


9.07 


Mass. 


16.56 


Mass. 


7.70 


N. Y. 


9.57 


Okla. 


8.32 


Pa. 


14.79 


Mo. 


6.13 


Ohio 


9.14 


Kan. 


6.79 


Conn. 


9.66 


111. 


5.55 


Conn. 


5.76 


Minn. 


6.02 


N. J. 


8.71 


Ky. 


5.48 


R. I. 


3.78 


Fla. 


4.74 


N. C. 


7.33 


Minn. 


4.54 


Iowa 


2.99 


Neb. 


4.61 


R. I. 


5.80 


Maine 


3.37 


N. H. 


2.83 


Wash. 


4.55 


Md. 


4.54 


Colo. 


3.01 


Minn. 


2.60 


Iowa 


4.52 


Va. 


3.91 


Kan. 


2.96 


Maine 


2.59 


N. Y. 


3.76 


Maine 


3.19 


Share of 
















ten leading 
















states in total 














net upward 
















shift 


80.47 




89.77 




64.46 




91.87 



Figures in italics are absolute net downward shifts. 
Source: Appendix Table A-6. 



Service Activities in the Regions, 1910-1950 

Services have been the most rapidly growing sector of the labor force during 
the twentieth century. Resource employment, the bulk of which is still agricul- 



Regional Adjustments within National Economy, 1910-1950 (2) / 257 

tural, has declined absolutely and relatively since 1910; manufacturing employ- 
ment has declined relatively since 1920, and since 1930 has barely held its own 
in absolute terms. In contrast, employment in services grew from around 14 
million to over 37 million between 1910 and 1950. The share of this sector in 
the total labor force rose by 71% over these years. In 1910 service activities 
employed about one-third of a total labor force of 38 million; in 1950, they 
employed nearly two-thirds of a total labor force of over 60 million. 

Changes in the regional distribution of the services labor force between 1910 
and 1930 appear to have been very slight, but since 1930 both the Southeast and 
the West have grown at the expense of the urban-industrial Northeast (Table 
107). In the West, all regions except the Plains have increased their shares 



Table 107. Regional distribution of services labor force, 1910, 1930, and 1950 



Region 


1910 


1930 


1950 


United States 


13,913,976 


23,012,448 


37,617,484 




100% 


100% 


100% 


New England 


8.43 


7.44 


6.27 


Middle Atlantic 


29.48 


29.00 


24.63 


Great Lakes 


20.15 


21.19 


18.95 


Southeast 


16.13 


15.45 


19.00 


Plains 


11.84 


9.94 


8.91 


Southwest 


4.72 


6.11 


7.89 


Mountain 


2.48 


2.03 


2.44 


Far West 


6.77 


8.84 


11.91 


Northeast 


58.06 


57.63 


49.85 


Southeast 


16.13 


15.45 


19.00 


West 


25.81 


26.92 


31.15 



Source: Appendix Table A-7. 



since 1930, but only the Far West and the Southwest have improved their rela- 
tive positions as compared with 1910. In the Northeast, the shares of the 
Middle Atlantic and Great Lakes regions have declined only since 1930, but 
New England has been losing relative to other regions since before the turn of 
the century. 

Table 108 compares the percentage growth in regional shares of all the 
United States service employments with the percentage growth in regional shares 
of all labor force and population since 1910. Because of the great rise in service 
employment in the Southeast, Southwest, and Far West, services as a proportion 
of the national total have declined more sharply than either labor force or 



- 13.20 


-15.31 


-25.62 


- 2.91 


- 0.42 


-16.45 


+ 1.66 


+ 8.83 


- 5.96 


- 6.31 


-15.38 


+ 17.79 


- 26.24 


-20.92 


-24.75 


+ 14.05 


+ 15.02 


+67.16 


+ 5.00 


- 1.34 


- 1.61 


+ 109.03 


+ 94.59 


+ 75.92 



258 / Regional Economic Development: 1870—1950 

population in all the Northeastern regions. Indeed, the Great Lakes region has 
registered increases in both total labor force and population but has nevertheless 
shown decline in services. The Southeast, on the other hand, has had a very 
marked increase in service employment in contrast to its shrinking shares of 
labor force and population. Within the West, the Plains region has declined 



Table 108. Percentage change in regional shares of population, total labor 
force, and services labor force, 1910-1950 



Services 
Region Population Labor force labor force 

New England 
Middle Atlantic 
Great Lakes 
Southeast 
Plains 
Southwest 
Mountain 
Far West 

Source: 17th U. S. Census, 1950, see supra, Table 1, note; Appendix Tables A-l, A-7. 



very sharply in all three categories, but more sharply in population and services 
than in total labor force. The Mountain region has declined moderately in labor 
force and service employments but has gained in population. Only the South- 
west and Far West have gained in all three categories. But whereas the South- 
west gained least in population and labor force and most in service employment, 
the Far West gained most in population and least in services. 

Considering the gross character of our tertiary sector (all non-resource and 
non-manufacturing employment), this analysis cannot be pressed too far, but 
it is instructive to conclude the discussion of employment shifts with a brief look 
at the shifts in service employment among the states. Table 109 summarizes the 
principal net upward and downward shifts among the states over the period 
1910-1950. The gains of the Far West were chiefly registered in California, 
involving more than 37% of the total net upward shift in the country. The 
Southwest gained mostly in Texas, but a further 6% of the total also accrued 
to Arizona, Oklahoma, and New Mexico. Florida is the great service state of the 
Southeast, but sizable net upward shifts also occurred in North Carolina and 
Virginia. Net downward shifts were most marked in the three great manu- 
facturing states of the Atlantic seaboard — New York, Pennsylvania, and Mas- 
sachusetts. The relative decline of the Plains states and of all the Great Lakes 
states except Michigan is no less remarkable. Michigan, in fact, is the only 
state in the urban-industrial Northeast to show any sizable balance of net 
upward shift of service employments over the last four decades. 



Regional Adjustments within National Economy, 1910-1950 (2) / 259 



Table 109. Principal net shifts in services labor force among states, 1910-1950 





Percentage net 






Percentage 


: net 






upward shift 






downward 


shift 




California 




37.29 


New York 






22.09 


Texas 




18.73 


Pennsylvania 




16.42 


Florida 




11.05 


Massachusetts 




12.30 


Michigan 




7.51 


Illinois 






9.60 


N. Carolina 




6.28 


Missouri 






7.87 


Virginia 




3.35 


Iowa 






4.52 


Arizona 




2.30 


Minnesota 






3.87 


Oklahoma 




1.85 


Ohio 






3.13 


New Mexico 




1.65 


Kansas 






2.37 


Oregon 




1.25 


Indiana 






2.30 



Source: Appendix Table A-7. 



These tendencies can be given clearer definition by considering certain major 
service components separately; namely, employment in transport and communi- 
cations, in trade and finance, in professional activities, and in personal and 
domestic service (Table 110). In the period before 1910, the first three of these 
service categories rose at a faster rate than service employments as a whole ; but 
only two of the four have maintained this distinction in the years since 1910, 
namely, trade and finance and professional employment. Transport and com- 
munications, which had run so far ahead of the others during the heyday of the 
westward movement, fell back into third place over ensuing decades. Personal 
and domestic services were unique in that, after 1910, they declined absolutely. 

There were, of course, regional differences in the distribution of these four 
categories after 1910, although the rate of redistribution for all regions was 



Table 110. Percentage growth in total labor force, services labor force, and 
labor force in four selected service industries, 1910-1950 



ALL LABOR FORCE 

ALL SERVICE LABOR FORCE 



+ 57.75 
+ 170.36 



Transport and Communications 
Trade and Finance 
Professional Services 
Personal and Domestic Services 



+ 65.61 
+ 244.84 
+ 180.99 
- 8.58 



Source: Appendix Tables Al and A7; 13th U. S. Census, 1910, Vol. IV, Occupational 
Statistics; 17th U. S. Census, 1950, Vol. II, State Volumes. 



260 / Regional Economic Development: 1870-1950 

lower than in the earlier period. Only three regions increased their shares of 
the total employment in all four categories: the Southeast, Southwest, and Far 
West (Table 111). 11 Apart from these three, the Great lakes region increased 
its share of transportation and communication workers, and the Mountain 
region gained very slightly in trade and finance. But both these regions declined 
relatively in the other three categories. The Middle Atlantic region lost rela- 
tively in all categories, but most heavily in personal and domestic services. New 



Table 111. Regional distribution of labor force in four selected service 
industries, 1910 and 1950 





Transport and 


Trad 


e and 


Professional 


Personal and 




communications 


finance 


services 


domestic 


services 


Region 


1910 


1950 


1910 


1950 


1910 


1950 


1910 


1950 


United States 


100% 


100% 


100% 


100% 


100% 


100% 


100% 


100% 


New England 


7.17 


5.31 


8.58 


6.36 


8.48 


7.23 


8.28 


5.49 


Middle Atlantic 


: 27.07 


25.98 


29.66 


25.21 


26.44 


24.77 


29.03 


23.05 


Great Lakes 


20.92 


21.14 


21.27 


20.31 


21.10 


19.83 


17.61 


16.48 


Southeast 


15.33 


17.00 


13.78 


17.26 


14.67 


17.64 


22.20 


26.80 


Plains 


13.19 


10.13 


12.75 


9.61 


13.98 


9.85 


10.08 


7.31 


Southwest 


5.10 


7.29 


4.94 


7.66 


5.41 


7.06 


4.75 


8.96 


Mountain 


3.45 


2.79 


2.21 


2.27 


2.79 


2.55 


2.09 


1.93 


Far West 


7.77 


10.36 


6.81 


11.32 


7.13 


11.07 


5.96 


9.98 



Source: 13th U. S. Census, 1910, Vol. IV, Occupational Statistics. 
Vol. II, State Volumes. 



17th U.S. Census, 1950, 



England had proportionately greater declines than the Middle Atlantic region 
in all four categories but its greatest losses were also in personal and domestic 
service. The Plains suffered quite heavy losses in all four categories; the great- 
est being in professional services. It is interesting that, since 1910, the three 
service "growth" regions — the Southeast, the Southwest, and the Far West — 
have all increased their shares of the generally declining personal and domestic 
services category. 

The percentage of the total labor force employed in each of the four selected 
service categories is shown in Table 112 for the eight regions and the nation as 
a whole. All of these categories except personal and domestic services were a 
larger proportion of the national labor force in 1950 than in 1910. In every 
region, the share of personal and domestic services in the regional labor force 
declined over these years, and the share of transportation and communication 

11 The Southwest and Far West, incidentally, were the only regions that increased their 
shares in all four categories over the entire period 1870-1950; New England and the 
Middle Atlantic were the only regions to experience relative losses in all four categories 
over the entire period. 



Regional Adjustments within National Economy, 1910-1950 (2) / 261 

workers was smaller in 1950 than in 1910 in four regions — the New England, 
Great Lakes, Mountain, and Far West regions. 12 

The uneven regional redistribution of these services is attested by the fact that 
by 1950 no region was above the national level for all four categories, while in 
1910 both the Middle Atlantic region and the Far West exceeded the national 
level. The New England and Great Lakes regions both exceeded the national 
level in three categories in 1910 but in only one category in 1950. The South- 
east and the Southwest fell short of the national level in all four categories in 
1910; by 1950, the Southeast was still below the national level in three cate- 
gories, but the Southwest exceeded it in all but one category. The Southeast was 
the only region in the country which fell short of the national level in trade and 
finance by any appreciable amount in 1950. It was in this category, moreover, 
that the rate of increase was greatest, both nationally and regionally, in the 
present century. 



Table 112. Labor force in four selected service industries as a percentage of 
the total labor force, by region, 1910 and 1950 









1910 






1950 






Trans- 








Trans- 










port & 


Trade 




Pers. 


port & 


Trade 




Pers. 


Region 


Comm. 


& Fin. 


Prof. 


& Dom. 


Comm. 


& Fin. 


Prof. 


& Dom. 


United States 


6.91 


9.47 


4.36 


9.88 


7.26 


20.70 


7.77 


5.80 


New England 


6.48 


10.64 


4.84 


10.71 


5.96 


20.36 


8.68 


4.92 


Middle Atlantic 


7.94 


11.92 


4.89 


12.17 


8.04 


22.25 


8.20 


5.69 


Great Lakes 


7.60 


10.59 


4.84 


9.15 


7.41 


20.32 


7.44 


4.61 


Southeast 


4.32 


5.33 


2.61 


8.95 


5.95 


17.24 


6.61 


7.49 


Plains 


7.82 


10.36 


5.23 


8.55 


7.97 


21.58 


8.29 


4.59 


Southwest 


5.69 


7.56 


3.81 


7.58 


7.42 


22.26 


7.70 


7.29 


Mountain 


10.67 


9.36 


5.44 


9.22 


9.16 


21.26 


8.97 


5.08 


Far West 


10.36 


12.43 


6.00 


11.36 


7.46 


23.23 


8.52 


5.74 



Figures in italics are above the national average. 

Source: 13th U. S. Census, 1910, Vol. IV, Occupational Statistics; 17th U. S. Census, 
1950, Vol. II; State Volumes. 



The measure of the Southeast's backwardness in this most rapidly expanding 
service sector can be seen in the fact that, although its share of trade and finance 
expanded far more than that of any other region — or, indeed, than that of the 
nation as a whole — it was still the furthest below the national level in 1950 
(Table 112). It is in this crucial trade and financial category, and to a lesser 

12 Of these four regions, however, only New England's proportion of transport and 

communications workers was smaller than the national proportion. On the other hand, 

the Southeast, which had enlarged its regional proportion in this category, was still below 
the national proportion in 1950. 



United States 


Northeast 


+ 118.6 


+ 91.4 


Southeast 


Plains 


+ 223.4 


+ 108.3 



262 / Regional Economic Development: 1870—1950 

Table 113. Percentage change in trade and finance shares of regional and total 
labor forces, 1910-1950 

Middle Atlantic Great Lakes 

+86.7 +91.9 

Southwest Mountain Far West 

+ 194.5 +127.1 +86.9 

Figures in italics are above the national average. 

Source: Same as Table 112. 



extent in professional employment, that the expanding service sector is most 
likely to grow in the immediate future. All regions of the country can expect to 
participate in the development, and the greatest relative gains will probably 
occur in the Southeast and Southwest. 

In view of the prolonged relative decline in resource activity and the more 
recent relative decline in manufacturing, the changing structure of service em- 
ployments, has from the standpoint of expanding population and labor force, 
become of crucial importance. Productivity per worker in services does not 
generally increase at rates comparable to those in the resource and manufactur- 
ing sectors; hence services are likely to continue for some time as the principal 
sector for labor force growth on the national level and in most of the regions. 

Within the service sector, moreover, shifts among the major categories of 
employment provide a revealing index to recent social-economic change. Since 
World War I, for example, the two categories of trade and finance and profes- 
sional services have overtaken transport and communications as leading forms 
of service growth. Completion of the railroad network before 1910 and chang- 
ing modes of transportation in the inter-war years help account for the slack- 
ening in this important category. 13 On the other hand, the increasing signifi- 
cance of consumption, reflected in the growth of nation-wide marketing, the 
proliferation of retail services, and the rise of insurance, savings, and credit 
services explains much of the buoyancy in trade and finance. In this same 
connection, the absolute decline in personal and domestic services registers the 
shrinking of the labor force available for domestic work, for there is every 
indication that personal services outside the home have greatly expanded during 
the last forty years. Finally, the upsurge in professional activity since World 
War I is the understandable concomitant of further developments in the urban- 
industrial organization of society, with its dependence on scientific, technical, 
and organizational "expertese." 



13 Part of the change, however, is explained by the difficulty of meaningful classification 
within the service sector. Should not the large numbers employed in servicing automobiles 
and their drivers, for example, be classified as transportation workers? 



YJ / Regional adjustments within the 
national economy, 1910-1950: 

(3) REGIONAL INDUSTRIAL STRUCTURES 
AND INCOME 



The changes that have occurred since 1910 in the structure and geographical 
distribution of productive activities within the national framework have had 
effects on the structure of activities in the regions themselves, which we shall 
now consider. By the early twentieth century, as we have seen, the farthest 
reaches of the United States had been brought into the national economy, and 
continued growth no longer involved the settlement and development of "empty" 
territories. For practical purposes, all parts of the country were accessible to 
one another. The completion of the Panama Canal and the introduction of 
truck-hauling services effectively reduced the transport costs that had for so 
long retarded the full development of the greater West. The resource potential 
of every region was finally at the disposal of a truly nation-wide system of 
production and marketing. The gradual decline of regional interests and wage 
differentials during the present century reflects this greater effective mobility 
of capital and labor. 

Thus, by 1910, the stage was set for a period of intensive interregional 
competition in which every type of productive activity must sooner or later 
meet the test of higher outputs at lower unit cost. 1 Changes in the industrial 
structures of the eight regions since 1910 reflect both this heightening of inter- 
regional competition and, more variously, the secular movement of labor out 
of resource activities and manufacture into services. 

The effects of the long-run secular change in the national labor force are 
summarized once again in Table 114. The period 1910-1950 was one in which 
all types of production except services declined in their shares of the total labor 

1 The progressive reduction of the freight factor in total costs of production, especially 
in regard to movements of raw materials, has enlarged the geographical area in which 
production is potentially located. While production has become more "foot-loose" in 
this sense, competition among sites for a given activity has generally been heightened. 

/ 263 



264 / Regional Economic Development: 1870-1950 

force. Agriculture had reached its peak share, of course, long before 1870 and 
the onset of nation-wide industrialization; fisheries had peaked about 1890 and 
forestry and logging at roughly the same time. Mining and manufactures 
achieved their largest relative shares of the labor force in 1920. Only the 
tertiary or services sector represented a larger share of the total labor force 
of the nation in 1950 than in 1910 (and there is no reason to suppose that, if 
current classifications are maintained, it will not continue to increase for a 
considerable time to come). These changing proportions of the national labor 
force provide a yardstick with which to measure and compare the levels of 
productive activities in the various regions. 



Table 114. Labor force, distribution by industry, 1910-1950. 



(per cent of total) 



Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Manufacturing Services 


1910 


32.46 


2.53 


0.45 


0.18 


27.92 36.46 


1920 


25.63 


2.62 


0.52 


0.13 


30.90 40.20 


1930 


21.44 


2.02 


0.36 


0.15 


28.90 47.13 


1940 


17.53 


1.84 


0.38 


0.12 


21.01 59.12 


1950 


11.56 


1.54 


0.36 


0.13 


23.92 62.4? 



Source: Appendix Table A1-A7. 



Industrial Structure of the New England Labor Force, 1910-1950 

The industrial structure of New England's labor force is shown in Table 115. 
The resource sectors as a whole have continued to decline in importance since 
1910. Mining, mostly quarrying, now employs only one-tenth of 1% of the 
regional labor force, and agriculture upwards of 3%. New England, in fact, 



Table 115. Distribution by industry of New England region labor force, 
1910-1950. 

(per cent of total) 



Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Manufacturing 


Services 


1910 


9.65 


0.31 


0.31 


0.43 


49.05 


40.25 


1920 


6.93 


0.15 


0.69 


0.28 


50.63 


41.32 


1930 


6.20 


0.19 


0.31 


0.33 


43.10 


49.87 


1940 


4.52 


0.14 


0.42 


0.25 


34.24 


60.43 


1950 


3.15 


0.13 


0.34 


031 


35.46 


60.61 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



Regional Adjustments within National Economy, 1910-1950 (3) / 265 

has very little activity left in the resources sector. It is worthy of note, however, 
that its proportion of workers in fisheries, though smaller in 1950 than in 1910, 
was above the national level throughout this period, as was the proportion in 
forestry during two of the decades under review. 

In spite of relative decline in most of its basic industries in recent years, 
New England is still a heavily concentrated manufacturing region. By 1930, 
however, its expanding service sector had surpassed its declining manufacturing 
sector. 

The differences between the industrial structures of northern and southern 
New England were somewhat less marked in 1950 than in the years before 1910. 
Although Connecticut and Rhode Island were still the most concentrated manu- 
facturing states in the region in 1950, New Hampshire had surpassed Massa- 
chusetts and risen to third place. Massachusetts, however, retained its priority 
in the service sector and in fisheries. Vermont was still by far the most agri- 
cultural state in the region in 1950, with more than 17% of its workers on 
farms; Maine had the second highest relative concentration of loggers in the 
country, being surpassed only by Oregon. 



Industrial Structure of the Middle Atlantic Labor Force, 1910-1950 

Table 116 shows the industrial structure of the Middle Atlantic labor force 
in the recent period. In 1950 this region had the highest concentration of the 
nation's services. The growth of commerce and finance in the metropolitan 
areas and the enormous expansion of government and related activities in 



Table 116. Distribution by industry of Middle Atlantic region labor force. 
1910-1950. 

(per cent of total) 



Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Manufacturing 


Services 


1910 


10.39 


3.94 


0.15 


0.16 


39.75 


45.61 


1920 


7.38 


3.45 


0.15 


0.09 


40.43 


48.50 


1930 


5.72 


2.65 


0.08 


0.09 


35.72 


55.74 


1940 


4.44 


1.97 


0.07 


0.07 


26.32 


67.13 


1950 


3.09 


1.48 


0.07 


0.07 


29.68 


65.61 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



and around the District of Columbia have reinforced the tendency for the 
tertiary activities to expand at a more rapid rate than the primary and secondary 
sectors. The Middle Atlantic region, in fact, has become the least agricultural 
region in the country. In all other resource activities, also, it has grown con- 



266 / Regional Economic Development: 1870-1950 

siderably less important, notwithstanding the great concentration of coal miners 
in Pennsylvania. Like New England, and indeed like advanced manufacturing 
areas elsewhere in the world, the Middle Atlantic region depends increasingly 
on the material resources of other less developed regions for its economic 
progress. 

As in 1910, New Jersey had the heaviest concentration of manufactures in 
this region in 1950 and Delaware the heaviest concentration of agricultural 
labor. Pennsylvania, of course, had the largest proportion of mine workers, 
and Maryland (together with the District of Columbia) the heaviest concen- 
tration of service workers, having recently surpassed New York. 



Industrial Structure of the Labor Force in the Great Lakes Region, 1910-1950 

The Great Lakes region's industrial structure is shown in Table 121. During 
the present century, the growth of its labor force has been based on manu- 
factures and services rather than resources; its resources have been in sharp 
decline since about 1930. Indeed, the Great Lakes region is unique among the 
more advanced industrial regions in having a slightly heavier concentration 
of manufactures in 1950 than in 1910, though manufacturing here has not 
regained its pre-depression peak and may even now be in relative decline. 
Michigan and Ohio have the greatest concentration of manufacturing labor in 
the region ; Wisconsin and Illinois have the lowest, yet both exceed the national 
level of manufactures by a very considerable margin. 



Table 117. Distribution by industry of Great Lakes region labor force, 1910- 
1950 

(per cent of total) 



Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Manufacturing 


Services 


1910 


25.16 


2.67 


0.31 


0.10 


33.19 


38.63 


1920 


18.72 


2.54 


0.41 


0.05 


37.36 


40.92 


1930 


14.36 


1.43 


0.19 


0.06 


35.72 


48.24 


1940 


12.51 


1.13 


0.13 


0.04 


28.88 


57.31 


1950 


8.36 


0.87 


0.10 


0.04 


33.43 


57.20 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



Until the 1930's, the Great Lakes region had a substantial agricultural sector, 
but since 1940 this has undergone a sharp relative decline, and by 1950 
agricultural workers comprised little more than 8% of the regional labor force. 
Wisconsin, however, had the heaviest concentration of farm labor in the entire 
urban-industrial Northeast in 1950 (17.9%), and Indiana had the third largest 

(n%). 



Regional Adjustments within National Economy, 1910-1950 (3) / 267 

The decline of mining in the region has been due mainly to the relative eclipse 
of iron and copper mining in Michigan. Though much of the fuel supply is 
still obtained locally, some of the mining areas must be included among the 
relatively depressed parts of the country. The forests of the region have also 
become a wasting asset; even Wisconsin is now well below the national level of 
forest employment. More surprising, the proportion of the regional labor force 
in service employment has also slipped below the national level in recent years; 
in 1950 Illinois had the region's largest concentration of service workers — 
61.7% — but remained just below the national proportion of 62.5%. Except for 
Michigan, services have grown at a lower rate than in the nation as a whole. 
It is in manufactures, therefore, that the Great Lakes region has shown the 
greatest relative strength in the 1910-1950 period. 2 



Industrial Structure of the Labor Force in the Southeast, 1910-1950 

As recently as 1930 the President of the United States referred to the South- 
east as "the Nation's No. 1 economic problem." At that time, while the region 
had only 9% of the national income, it had 21% of the nation's population and 



Table 118. Distribution by industry of Southeast region labor force, 1910-1950 













(per cent 


of total) 


Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Manufacturing 


Services 


1910 


58.93 


1.73 


0.65 


0.23 


14.46 


24.00 


1920 


49.49 


2.72 


0.67 


0.20 


18.38 


28.54 


1930 


42.21 


2.52 


0.57 


0.28 


19.78 


34.64 


1940 


33.41 


2.60 


0.77 


0.23 


15.40 


47.59 


1950 


21.35 


2.60 


0.76 


0.24 


17.79 


57.26 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



the highest birth rate in the country. 3 Table 118 shows the changing industrial 
structure of the Southeast since 1910. Compared with the earlier period, there 
has been a significant industrial expansion over the last four decades. However, 
manufactures have risen only very slowly, though the Southeast is the only 

~ In the late nineteenth century the Great Lakes region became the largest producer of 
silage corn and whole milk. It is still rich in coal and oil resources and formerly had the 
greatest regional production of iron and copper ore. The region has the cheapest and 
most efficient system of natural inland waterways; it produces more automobiles, processes 
more rubber, makes more farm equipment and, along with the Plains region, packs more 
meat than any other part of the country. 

3 R. B. Vance, The South' 's Place in the Nation, (Public Affairs Pamphlet, No. 6, 
Revised 1941), pp. 8, 14; H. W. Odum, Southern Regions of the United States (Chapel 
Hill: University of North Carolina Press, 1936), p. 51. 



268 Regional Economic Development: 1870-1950 

region except the Great Lakes to have a higher proportion of its labor force in 
manufactures in 1950 than in 1910. In 1950. however. Southeastern manufac- 
turing had not regained the structural importance it had achieved in 1930. It is 
perhaps too early to judge whether this crucial sector is now in relative decline, 
as it appears to be everywhere else in the country. 

It is clear that the Southeast is still a major resource region: it is the only 
region in the country to be above the national level in all four primary resource 
sectors. The manpower in agriculture has shown a steady decline, especially 
since 1930. while mining has gathered strength. Between 1940 and 1950 the 
Southeast displaced the Middle Atlantic region as the third most concentrated 
mining region in the country. Its forests and fisheries are major suppliers for 
the nation, and employment in both sectors appears buoyant. 

Within the region itself, the range of industrial structures exhibited by the 
states is very great, yet no single state presents a structural profile which resem- 
bles that of the nation as a whole. The Carolinas and Georgia, for example, 
come closest to the national proportion of manufacturing employment, but their 
agricultural sectors still comprise from one-fifth to one-quarter of their entire 
labor forces. In 1950. eight states in the Southeast had larger agricultural than 
manufacturing sectors. Mississippi was at the top of the list with 40.8% of its 
labor force in agriculture and only 10.6% in manufactures: its service sector. 
46.6%. was the lowest in the region and in the nation. 

Five Southeastern states — West Virginia. Kentucky. Louisiana. Alabama, and 
Virginia — are above the national level in mining employment. West \ irginia. 
in fact, has the greatest concentration of mining employment of any state in the 
Lnion. 20.3 % ; its closest rivals are Wyoming with 7.2% and. in the Southeast. 
Kentucky with 6.8%. West Virginia has a greater share of its total labor force 
in mining than in either agriculture (9.4%) or manufactures (17.2%). In view 
of the secular decline in coal mining, this is by no means a favorable situation. 

Georgia. Arkansas, and Mississippi have the largest concentrations of logging 
employment, ranging from 1.2% to 1.3% of their respective labor forces, while 
Kentucky has the lowest. 0.3%. Louisiana and Florida have the largest propor- 
tions in fisheries. 0.8% and 0.6%. and West Virginia the lowest, only one-tenth 
of 1%. Florida has by far the largest concentration of service employment. 
77.7%. a proportion which is exceeded in only one other state. Nevada, and the 
District of Columbia. 



Industrial Structure of the Labor Force in the Plains Region. 1910-1950 

The Plains region presents a somewhat different kind of economic growth 
problem. Its population density is rising very slowly, for its decennial increases 
in population have been the lowest in the nation since 1890: its share of national 
population has declined bv more than 26% since 1910 and its share of the 
labor force by 20%. This is a great farming region. By 1950 the Plains region 
had an even higher concentration of agricultural workers in its labor force than 
the Southeast (Table 119). 



Regional Adjustments within National Economy, 1910-1950 (3) / 269 

The region's contribution to the nation's iron ore supply since the 1890's has 
been monumental, but employment in mining (chiefly in Minnesota) has, with 
minor interruption, been in relative decline since 1920. Meanwhile employment 
in forests and fisheries has also gone into eclipse. Manufactures reached their 
peak share of the regional labor force in 1920 and declined by almost one-half 
during the depression decade. There has been partial recovery since 1940 in 



Table 119. Distribution by industry of Plains region labor force, 1910-1950 













(per 


cent 


of total) 


Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Manufacturing 


Services 


1910 


40.85 


1.82 


0.29 


0.04 


19.99 




37.07 


1920 


36.35 


1.62 


0.44 


0.03 


21.04 




40.52 


1930 


33.59 


1.05 


0.18 


0.04 


19.85 




45.29 


1940 


30.46 


0.95 


0.15 


0.02 


11.26 




57.16 


1950 


23.84 


0.86 


0.14 


0.03 


14.75 




60.38 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



some of the eastern Plains states, but it has not been adequate to bring the region 
much closer to parity in manufacturing employment with the rest of the nation. 
Since 1920, in fact, the Southeast has made a much better showing in manu- 
factures than the Plains region and has experienced a greater percentage 
increase in its service sector. 

By 1950 Missouri was the only Plains state which almost approached the 
national proportion of manufacturing employment, and the only one with a 
manufacturing concentration greater than its agricultural concentration. North 
Dakota has the lowest proportion of manufacturing employment in the nation 
(2.8%) ; South Dakota, the third lowest (4.6%). North Dakota has the nation's 
greatest concentration of agricultural labor — 43.2%; and South Dakota, with 
39%, has the third highest, following Mississippi in the Southeast. Minnesota, 
with one-third of 1% of its labor force in logging is the leading forest employer 
in the region. And Minnesota with 1.3% and South Dakota with 1.08%, have 
the largest proportions of mining employment; Nebraska with only two-tenths 
of 1% is the lowest. Kansas and Nebraska have the highest concentrations of 
service employment, 64.1% and 62.3% respectively, while North Dakota has the 
lowest, 53.6%. Clearly, the saving grace of the Plains region is the fact that its 
population growth is slight. It has escaped the threat of relative impoverish- 
ment through the willingness of its people to follow employment opportunities to 
other parts of the country. Relative net downward population shifts in the seven 
Plains states between 1910 and 1950 amounted to more than one-third of the 
total net downward shift for the nation. 4 

4 Missouri, Iowa, and Kansas alone accounted for 22.27% of the total net downward 
shift in population between 1910 and 1950 (See Table 82). 



270 / Regional Economic Development: 1870-1950 



Industrial Structure of the Labor Force in the Southwest, 1910-1950 

Table 120 shows the industrial structure of the Southwest's labor force since 
1910. The region has remained primarily a resource region with important 
segments of the nation's agriculture, mining, and, to a lesser extent, forestry 



Table 120. Distribution by industry of Southwest region labor force, 1910-1950 

(per cent of total) 



Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Manufacturing 


Services 


1910 


57.67 


1.77 


0.36 


0.05 


12.37 


27.78 


1920 


44.88 


3.49 


0.30 


0.03 


15.62 


35.68 


1930 


37.24 


2.90 


0.27 


0.05 


17.45 


42.09 


1940 


29.15 


3.47 


0.28 


0.06 


8.11 


58.93 


1950 


16.01 


3.52 


0.22 


0.08 


11.01 


69.16 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



carried on within its bounds. Its fishing resources are as yet very slight, but the 
Gulf ports of Texas are contributing more to the nation's fisheries each decade. 
Since 1930, the region has rapidly increased the relative size of its service 
sector, and by 1950 only the Far West had a proportionately larger body of 
service workers. Manufactures, which were still growing relatively through 
1930, declined sharply in the depression years, but their relative importance has 
increased again since 1940. Nevertheless, the Southwest had the second lowest 
concentration of manufactures in the country in 1950. 

The heaviest manufacturing concentration in this region has been achieved 
in recent decades by Texas, which had about 12% of its labor force in manufac- 
tures by 1950. 5 New Mexico is now the least affected by industrialism, with only 
5.1% in manufactures. Oklahoma has the highest proportion of farm labor in 
the region — 19.6% — and Arizona the lowest — 14.1%. Oklahoma also has the 
highest relative concentration of workers engaged in mineral extraction — almost 
5% in 1950 — while Texas, somewhat surprisingly, has the lowest — 3%. The 
regions' forest resources are very small. Arizona has the largest proportion of 
forest labor, but that amounts to only three-tenths of 1%. All four Southwestern 
states have high proportions of service employment, ranging from 73.8% in 
Arizona to 66.1% in Oklahoma. In absolute terms, of course, Texas dominates 
the Southwest, and its enormous endowments of agricultural and mineral wealth 
have served to enrich not only the region but the entire nation. 

5 In 1910 Arizona had the highest proportion of manufacturing, 20.8%, and Texas the 
lowest, 11.8%. 



Regional Adjustments within National Economy, 1910-1950 (3) / 271 



Industrial Structure of the Labor Force in the Mountain Region 1910-1950 

The industrial structure of productive activities in the Mountain region 
between 1910 and 1950 is presented in Table 121. This region also remains 
primarily a resource region. Throughout most of the period it had the heaviest 
relative concentration of mineral workers in the country; by 1950, however, it 
was surpassed in this respect by the more populous Southwest. Since 1920 its 
proportion of agricultural employment has been above the national average. In 



Table 121. Distribution by industry of Mountain region labor force, 1910-1950 

(per cent of total) 



Year 


Agriculture 


Mining 


Forestry 


Fisheries 


Mi 


anufacturing 


Services 


1910 


29.93 


8.47 


0.67 


0.02 




20.48 


40.43 


1920 


32.85 


6.65 


1.07 


0.01 




18.72 


40.70 


1930 


30.95 


5.25 


0.98 


0.02 




18.12 


44.68 


1940 


25.12 


5.10 


0.61 


0.03 




7.64 


61.50 


1950 


17.72 


3.44 


0.50 


0.04 




9.46 


68.84 



Figures in italics are above the national average. 
Source: Appendix Tables A1-A7. 



logging employment, the peak share of the national total was reached in 1930, 
but the peak share in the regional labor force had been reached a decade before 
this. Its manufacturing employment reached a peak in 1910, somewhat earlier 
than in other new regions of the West. After 1930, the most rapidly growing 
employment sector was services, which surpassed the national level throughout 
most of the period. 

Utah and Colorado have the heaviest concentrations of manufactures in the 
region (about 11% each) and Wyoming the lowest (5.2%). Idaho has re- 
mained the most agricultural state in the region (25.6%), closely followed by 
Montana (23.7%); Utah is now the least agricultural (11.8%). In mineral 
employment, Wyoming leads the region with 7.2% and Colorado is in the lowest 
position with only 2.0%. In forestry employment, Idaho has the heaviest con- 
centration, amounting in 1950 to 1.4%; Montana is in second place with 0.8%. 
Utah and Colorado lead in services activities, their service employments in 1950 
amounting to 71.6% and 72.3% of their respective labor forces. Idaho has the 
lowest regional share of services, 62.9%. The economic structure of this region 
suggests a rather unique resources-services base of economic activities whose 
progress is closely tied to developments in the major industrial centers of the 
country and to federal government policy. 



272 / Regional Economic Development: 1870-1950 



Industrial Structure of the Labor Force in the Far West, 1910-1950 

The promise for growth shown before 1910 by this great "empire apart" has, 
on the whole, been fulfilled in the intervening years. With every decade since 
1910 the Far West has absorbed a greater share of the nation's population. Its 
only rival has been the Southwest, where the rate of population increase has 
nevertheless been much smaller. The resulting density is still much below that 
of the urban-industrial Northeast or the largely rural-agricultural Southeast, 
but since 1940 it has exceeded that of the Plains by a growing margin. In the 
single decade 1940-50 the population of the Far West increased by almost 30%. 



Table 122. Distribution by industry of Far West region labor force, 1910-1950 

(per cent of total) 



Year 


Agriculture 


Mining 


Forestry 


Fisheries 


M; 


anufacturing 


Services 


1910 


20.10 


2.84 


2.04 


0.47 




26.95 


47.60 


1920 


18.33 


1.71 


1.85 


0.42 




28.89 


48.80 


1930 


14.56 


1.45 


1.41 


0.35 




25.99 


56.24 


1940 


11.51 


1.52 


1.18 


0.26 




14.83 


70.70 


1950 


7.44 


0.65 


0.99 


0.24 




16.85 


73.83 



Figures in italics are above the national average. 
Source : Appendix Tables A1-A7. 



Table 122 shows the changing industrial structure of this rapid growth region. 
The Far West has the smallest proportion of agricultural workers in its labor 
force anywhere outside the New England and Middle Atlantic regions; and its 
concentration of mineral activities has been below the national level since 1910. 
Its proportion of fishing employment still exceeded the national level in 1950, 
but it had fallen throughout the period. Forestry is the only primary resource 
sector in which the Far West retained national priority. 6 

Between 1930 and 1950 the Far West's labor force grew from about 3,500,000 
to 6,000,000 or from about 7.4% of the national labor force to 10%. During 
the same two decades the combined resource sector fell from 17.8% to 9.3%, 
and the combined manufacturing-service component rose from 82.2% to 90.7%. 
It was services, however, more than manufactures, that accounted for this rise 
in the industrial component. Over these years the proportion of the total 
regional labor force employed in services increased from 56.2% to 73.8%. 
Services, therefore, absorbed not only the bulk of the displaced primary workers 

6 Annual cut still exceeds growth; though this holds true for most logging areas, it is 
especially pronounced in the Far West, which contains the greatest share of the nation's 
forest resources. The immense construction needs of the region in recent decades and the 
availability of new timber have resulted in the virtual destruction of whole forests. 



Regional Adjustments within National Economy, 1910-1950 (3) / 273 

but secondary workers as well, and no doubt the larger portion of in-migrants. 

The somewhat faster rate of increase in the region's population than in its 
labor force during this period reflects the large proportion of persons under 
and over employment age. Thus the rising numbers of service workers were 
partly employed in serving a population which was not itself fully employed. 
Such important activities as trade and finance, the professions, education, 
recreation and entertainment, and, not least, construction, experienced con- 
siderable impetus from the investment of capital which had been accumulated 
outside the Far West itself. 

The small absolute decline in resource employment between 1930 and 1950, 
and a somewhat larger absolute decline in manufacturing employment, appear 
as very large relative declines in these sectors since 1930 (Table 122) when 
set against the absolute growth of the Far Western labor force in general and 
the service sector in particular. Whether this rate of growth in the service 
sector can be sustained is quite another matter; whether it is something peculiar 
to the spectacular growth of the Far West during the historical period 1930- 
1950 is something which can only be determined by future developments. 7 

In 1950 California had much the heaviest concentration of manufacturing 
employment within the Far West, 17%, and Nevada the lightest, only 4.6%. 
Oregon had the heaviest local concentrations of both agricultural and forest 
employment, 11.4% and 4.5% respectively. In fact, by 1950, Oregon had the 
heaviest relative concentration of loggers and woodsmen of any state in the 
country and the largest in absolute terms as well. Because of the very small 
size of its labor force, Nevada had the largest proportion of miners of any 
Far Western state — 4.7% — but in absolute terms, of course, they numbered 
less than one-tenth the mineral labor force in California. For the same reason, 
Nevada also had the heaviest concentration of service activities in the nation, 
81.5% — exceeded only by the special case of the District of Columbia; yet, 
compared with California or any other Far Western state, the actual number 
of service workers in Nevada was always insignificant. In relative terms, 
priority in the region's fisheries had by 1950 passed to Washington (0.54% 
compared with California's 0.18%), but in absolute terms the 8,000 fishermen 
of California outnumbered those of Washington or any other state in the Union. 

The enormous growth of the California labor force is one of the most striking 
phenomena in recent economic history. In 1910 California was only tenth 
among the states in the number of its workers; in 1950 it was surpassed only 
by New York itself. From less than a quarter of a million workers in 1870 
it had risen to almost four and a half millions at the mid-century. 

7 In many ways the Far West is the most favored region in the country. Rainfall reaches 
60 inches in some localities, while in arid areas irrigation projects permit a growing 
season up to 8 months. It is a region of profitable agricultural specialties, though water 
supply may prove to be a critical limitation. It is the largest lumber producer and has 
the richest fisheries; it is the leading gold producer and, more important, the second 
largest oil region after Texas-Oklahoma. Industrialization is based on its oil and water 
power resources. The region has recently become a center for ships, aircraft, aluminum, 
and scientific research. The rapidly growing population makes the region a major market 
for utilities and consumption goods. 



274 / Regional Economic Development: 1870-1950 



Income Changes, 1920-1950 

The structural adjustments in population and labor force described above 
were accompanied by significant changes in personal income payments among 
the various regions. These are analyzed in some detail in Part V of this book. 
However, to round off the descriptive picture of regional economic development 
in recent decades, we here present summary data on income changes during 
the period 1920-1950. 

Changes in the relative position of regions with regard to total personal 
income have characterized the entire period. Between 1920 and 1930, for 
example, a significant relative decline in total personal income occurred in four 
regions: the Mountain, Southeast, Southwest, and Plains regions, i.e. in the 
regions with the highest relative proportions of their labor forces in agriculture. 
In the Far West and Middle Atlantic regions, on the other hand, total personal 
income was higher in 1930 than in 1920. Over the depression decade, 1930-40, 
an essentially reverse change took place. During this interval, all of the agri- 
cultural regions except the Plains improved their relative positions, while two 
of the urban-industrial regions, the Middle Atlantic and New England, ex- 
perienced some decline. A similar pattern obtained for the period 1940-1950. 
Because of the unusual agricultural prosperity in 1950, all of the farm regions 
had a better income standing in this year than in 1940. 

Table 123 shows the effect of these varying regional rates of change on the 
shares of total personal income. 

The net shifts in total personal income among the states over the 1920-1950 
period as a whole are shown in Figure 41, and the principal upward and down- 
ward shifts are summarized in Table 124. In the Far West, every state registered 



Table 123. Regional distribution of total personal income in the United States, 
1920-1950 

(U. S. income in millions) 



Region 


1920* 


1930 


1940 


1950 


United States 


$69,276 


$76,780 


$78,522 


$225,473 




100% 


100% 


100% 


100% 


New England 


8.75 


8.58 


8.15 


6.73 


Middle Atlantic** 


30.20 


33.32 


30.50 


26.37 


Great Lakes 


22.21 


22.57 


22.69 


22.51 


Southeast 


13.02 


11.15 


13.23 


15.16 


Plains 


10.27 


8.86 


8.30 


8.81 


Southwest 


5.69 


4.75 


5.21 


6.50 


Mountain 


2.47 


1.93 


2.03 


2.22 


Far West 


7.39 


8.84 


9.89 


11.70 



* 1919-21 average. 

** 1920— D.C. omitted; 1930-50— D.C. included. 

Source: Same as Table 8. 



Regional Adjustments within National Economy, 1910-1950 (3) / 275 




Absolute upward shift 
II Relative downward shift 



Figure 41. Net Shift in Total Personal Income, 1920-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 14.7. Total net shift as % of 1920 personal income: 32.7. 



some net upward shift of personal income, and California alone accounted for 
more than a third of the national upward shift. In the Southeast, Florida, 
North Carolina, and Virginia made the greatest relative gains, with Tennessee, 
Louisiana, and Alabama showing less sizable increases. Michigan was the only 
state in the Great Lakes region that made any notable relative gain; indeed in 



Table 124. 
1950 



Principal net shifts in total personal income among states, 1920- 








Percentage 


net 




Percentage 


net 






upward shift 




downward 


shift 




California 






37.85 


New York 




28.04 


Michigan 






10.09 


Pennsylvania 




19.76 


Florida 






9.99 


Massachusetts 




15.36 


Texas 






9.97 


Illinois 




6.08 


N. Carolina 






5.20 


Oklahoma 




3.48 


Virginia 






3.93 


Kansas 




3.15 


Washington 






2.70 


Missouri 




3.14 


Oregon 






2.50 


Iowa 




2.56 


Tennessee 






2.48 


Maine 




1.84 


Indiana 






2.19 


Nebraska 




1.68 



Source: Same as Table 8. 



276 / Regional Economic Development: 1870-1950 

all the other Great Lakes states except Indiana there were relative net down- 
ward shifts, especially in Illinois. All New England states except Connecticut 
had relative downward shifts but only Massachusetts showed very serious 
relative decline. In the Middle Atlantic region, New York and Pennsylvania 
together contributed more than 47% of the total net downward shift in personal 
income; New Jersey, Maryland, and Delaware all registered net upward shifts. 
The Plains region was the only region in which every state suffered relative 
losses; the total for the region was 13.9%. In the Mountain region, only Utah 
registered a slight relative net upward shift; all other Mountain states under- 
went relative decline. 

Table 125 compares real personal income per capita in the eight regions for 
the years between 1920 and 1950 with the national average for these years. 
Throughout most of this period four urban-industrial regions rank above the 
national average and four rural-agricultural regions rank below. After 1920, 
when the Mountain region falls below the countryside level for the first time, 
the rank order of the relatively poorer states does not change, though after 
1930 all converge upwards toward the national average. The rank order of 
the richer urban-industrial regions, however, is subject to change in almost 
every decade, though after 1930 all converge downwards toward the national 
average. 



Table 125. Rank order of regions by personal income per capita,* 1920-1950 

(1929 dollars) 





1920 




1930 




1940 




1950 


F.W. 
M.A. 

N.E. 
G. L. 


780.6 
772.5 
719.3 
623.5 
590.7 


M.A. 
F.W. 

N.E. 
G.L. 


912.9 
837.0 
826.7 
701.6 


M.A. 
F.W. 

N.E. 
G.L. 


965.8 
959.7 
925.4 
815.4 


F.W. 
M.A. 
G.L. 

N.E. 


1,279.4 
1,255.2 
1,183.9 
1,160.4 


Mt. 


U. S. 


640.0 


U. S. 


727.4 


U. S. 

Mt. 

PL 

S.W. 

S.E. 


1,062.7 


U. S. 


578.0 


Mt. 
PL 
S.W. 
S.E. 


551.8 
523.1 
411.3 
321.0 


Mt. 
PL 
S.W. 
S.E. 


649.2 
590.5 
511.0 
419.3 


1,017.8 


PL 

S.W. 

S.E. 


501.0 
466.2 
326.2 


1,003.6 
916.6 
719.2 



* Weighted averages of state incomes per capita. 
Source: Same as Table 9. 



Relationship of Personal Income and Population 



A comparison of regional income shares with regional population shares 
for the years between 1870 and 1950 throws some light on changing levels of 
living in the various parts of the country during the period of rapid national 



Regional Adjustments within National Economy, 1910-1950 (3) / 111 

growth after 1870. Table 126 shows the year in which each region reached 
its peak share of the nation's total population and the year in which it reached 
its peak share of total personal income. It also describes the trend of per capita 
personal income in each region relative to the national average level since 1870. 
The significant feature of this trend is the great rise in real per capita income 
throughout the nation and the reduction in the disparity between regional 
average levels of living. Since 1920, the convergence in per capita income has 
become even more marked than in the period before that date, in spite of the 
transformations brought about by alternations of prosperity and depression. 

Some interesting connections between the growth of population and total 
personal income are reflected in the changing levels of per capita income in the 
various regions. Three of the eight regions reach their peak shares of both 
total population and personal income before the beginning of our period in 
1870: the New England, Middle Atlantic, and Southeast regions. 8 As these 
three regions had all been settled and partly developed during the eighteenth 
and first half of the nineteenth centuries, the development of newer regions to 
the west could be expected to be at the expense of their relative shares. The 
five other regions, however, reach or approach their peak shares of total popu- 
lation and personal income after 1870 in the following order: the Great Lakes, 
the Plains, the Mountain, the Southwest, and the Far West. The last two, in 
fact, appeared to be still enlarging their shares of the two national totals at the 
expense of the rest of the country as recently as 1950. It follows from what 
we have already noted about convergence in regional levels of per capita 
personal income that the redistribution of population and total personal income 
has proceeded at a much slower pace since World War I than before. 

The experience of the four Western regions (and of the Great Lakes region 
also) indicates that both population and total personal income shares grow 
very rapidly during the initial phases of settlement but that the rate of growth 
tends to slacken after a certain period has elapsed. As a result of this rapid 
initial growth, per capita personal income levels in the four Western regions 
were converging towards the national average by 1900. Unusually rapid popu- 
lation growth would seem to cease in each case at about the time when per 
capita personal income comes closest to the national average level — in other 
words, when total personal income ceases to maintain its earlier rapid rate of 
growth or when population has been increasing at a more rapid rate than total 
income. 

The Great Lakes region's share of national population peaked around 1870; 
over the ensuing decades the rate of population growth and the share of total 
population slowly declined. The share of total personal income continued to 

8 The fact that the Southeast has recently surpassed its 1880 share of total personal 
income once again does not belie the fact that its shares of both income and population 
peaked before the Civil War. There is reason to believe that the Southeast's share of 
total personal income had fallen drastically from pre-war proportions by 1880; see R. A. 
Easterlin, "Interregional Differences in Per Capita Income, Population, and Total Income, 
United States, 1840-1950," National Bureau of Economic Research, Conference on Research 
in Income and Wealth, Sept. 4-5, 1957 (mimeo). 



278 / Regional Economic Development: 1870-1950 



Table 126. Regional peaks of population and total personal income shares, and 
trend of regional relative to national per capita income in the United States 
since 1870 



Region 



Peak year for 

population 

relative to 

that of the 

nation 



Peak year for 

total personal 

income relative 

to that of the 

nation 



Trend of 

per capita 

income relative 

to that of the 

nation 



New England 



before 1870 



Middle Atlantic before 1870 



Great Lakes 



Southeast 



ca. 1870 



before 1870 



Plains 



Southwest 



Mountain 



Far West 



ca. 1890 



before 1880 



before 1880 



ca. 1880 



Only since 1940 
has it risen 
above 1880 
relative level 



ca. 1900 



1950 still rising 1950 still rising 



ca. 1920 



ca. 1900 



1950 still rising 1950 still rising 



Always above U. S. Rising be- 
fore 1880 but converging slowly 
downward toward U.S. there- 
after. 

Always above U.S. but converg- 
ing downward toward national 
level after 1900. 

Converges from below U.S. level 
about 1880, diverges above 
thereafter until about 1930. 

Below U.S. always, but con- 
verges upward slightly after 
1900. Diverges downward 
again after 1920 until late 
1930's when it converges once 
more. 

Always below U.S. but converges 
from below before 1900. Di- 
verging downward thereafter 
until about 1940. 

Always below U.S. but converg- 
ing slowly until 1920; diverg- 
ing downward until 1930, con- 
verging upward since 1940. 

Above U.S. before 1920 but con- 
verging downward sharply. 
Diverges below U.S. until 
1930's, converging upward 
again after 1940. 

Always above U.S., but converg- 
ing slowly downward to na- 
tional level throughout period, 
except for rapid upward diver- 
gence during 1930's. 



grow, however, and the level of per capita income surpassed the national 
average in all the years for which there are income data. In or around 1880, 
the share of total personal income was at its peak; thereafter it also declined, 
although more slowly than the share of total population. For this reason per 
capita personal income in the Great Lakes region continued to rise as compared 



Regional Adjustments within National Economy, 1910-1950 (3) / 279 

with the national average. Between 1900 and 1920 the region moved into 
fourth rank order of regional per capita income, behind the other three leading 
urban-industrial regions. 

The Plains region reached its peak share of national population around 1890; 
its rate of population growth and its share of total population then fell off quite 
rapidly. Total personal income reached its peak share a decade later, about 
1900, and thereafter sharply declined, somewhat more sharply in fact than the 
population share. As a consequence, per capita personal income in the Plains 
region came into closest convergence with the country-wide average in 1900, 
after which it continued to fall through 1920. 9 

In contrast, in the more sparsely populated agricultural Southwest, there was 
a gradual convergence on the national level throughout the 1870-1950 period. 
The rate of population growth began to slacken as early as 1880, but it 
nevertheless continued to be higher than the rate for the nation as a whole; 
hence the share of total population also continued to grow, although at a 
markedly slower pace after 1910. The region's share of total personal income 
also rose down through 1920 and, although the rate of increase was slower 
after 1900 than before, it stayed ahead of population growth at all times. 
Thus, the level of per capita income in the Southwest had a greater relative 
rise than that of the Plains region, even though in absolute terms it remained 
somewhat below the Plains level. 

The experience of the Mountain region differs somewhat from that of any 
other region although the main trend broadly parallels that of the Plains. As in 
the Southwest, however, the rate of population growth, despite some fluctuations, 
was somewhat higher than the national average rate throughout the years from 
1870 to 1920. The region reached its peak share in the national population 
around 1920, at which time it was still the most sparsely populated area in the 
country. The growth of total personal income, however, had meanwhile not 
sufficed to raise the share of the national total beyond the level of 1900; in fact, 
the share had declined some 2% by 1920. Hence, while the population share 
increased after 1900, the personal income share fell somewhat, and the effect 
on per capita personal income was decline. Personal income per capita in the 
Mountain region had been among the highest in the nation around 1880-1900. 
After 1910 it converged rapidly downward toward the national average, and 
after 1920 it was below the national average for the rest of the period. 

The Far West alone among the newer Western regions maintained a very 
high level of per capita income throughout the entire 1870-1950 period. 
Though not by Eastern standards a densely populated region, the Far West is 
heavily populated. From 1870 to 1890 it had a heavy rate of population in- 
crease, which slackened only during the depression years of the nineties; and 
after 1900, its decennial rate of increase was, without exception, considerably 
greater than that of any other region. But through most of the period since 
1900 its total personal income has maintained a fast rate of growth. Hence, 
despite its sustained population growth, the Far West has had the highest 

9 The Plains did not arrest its relative decline in per capita personal income much 
before 1940. 



280 / Regional Economic Development: 1870-1950 

level of per capita income of any region in the country during the present 
century except for a brief interval in the inter-World War years when it was 
surpassed by the Middle Atlantic region. Nevertheless, its per capita income 
level has been slowly converging toward the national level in recent years. 

The share of New England in the nation's population declined steadily 
throughout the 1870-1950 period. But that of the Middle Atlantic region, though 
it declined between 1870 and 1890. turned upward again after 1900 and con- 
tinued to grow through 1940. In both these regions, it is likely that per capita 
income levels rose high above the national average over the third quarter of 
the nineteenth century. 10 After 1880. however, they gravitated slowly down- 
ward toward the national level. But the growth of total income in the Middle 
Atlantic region was such that the fall in its share of the national total was very 
slight, and after 1900 almost negligible. As a result, this region rose from 
fourth to third place in rank order of regional per capita income between 1880 
and 1900 and from third to second place between 1900 and 1920. From the 
late twenties until after 1940. it ranked first among the regions in per capita 
income l Table 125). In New England, on the other hand, the shares both of 
population and of total personal income declined continuously after 1900. The 
rate of decline was always greater, however, for population than for income, 
while both were greater in the period before 1900 than in the years between 
1900 and 1920. Between 1880 and 1900 Xew England fell from third to fourth 
place in rank order of regional per capita income. Between 1900 and 1920 it 
rose again to third place owing to the rapid decline of the Mountain region. 
But by 1950 it had been surpassed by the Great Lakes region and was once 
more fourth in rank. 

Developments in the Southeast differ from those in any other region. The 
trends here are complicated somewhat by the fact that, during the period under 
review, the states in the western part of the region were growing while those 
on the South Atlantic seaboard were in continuous decline. On balance, how- 
ever, the Southeast's share of total personal income declined between 1880 and 
1900. but much of the lost ground was recovered before 1920 by gains in 
individual states. The population share declined at about the same rate in 
both periods. The declining population share, accompanied by a rising share in 
total personal income after 1900. brought the Southeast's level of per capita 
personal income back towards the average for the nation as a whole. In 
absolute levels of per capita income, the region was only slightly better off in 
1920 than in 1880: the rise over these years was from 50% to 56% of the 
national average. Between 1920 and 1950. however, the Southeast's share 
rose significantly to 68% of the national average. 

Thus, with minor qualifications, per capita incomes everywhere seem to have 
converged towards the national level during the opening decades of the 
twentieth century. And with the possible exception of the Southeast, this trend 
continued through the entire period 1880-1950. The newer Western regions 
and the Great Lakes region reveal strikingly similar patterns of association 

10 R. A. Easterlin. "Interregional Differences in Per Capita Income. Population, and Total 
Income, United States. 1840-1950," he. cit., passim. 



Regional Adjustments within National Economy, 1910-1950 (3) / 281 

between population and total personal income; in all these regions the rapid 
growth of population and total personal income shares is accompanied by a 
marked convergence of per capita personal income levels towards the national 
average. When the rate of growth in regional income shares begins to slacken, 
as it always does sooner or later, the trend towards parity in regional levels 
of per capita income is also gradually arrested. The more urban-industrial 
regions seem to hold their levels of per capita income above the national average, 
while the more agricultural, less urban-industrial regions seem to settle at 
below-average levels. In either case, rapid growth apparently ceases at or 
around the time when per capita personal income levels approach close to the 
national average. Moreover, if per capita income levels are moving down 
toward the national average, as for example in the Mountain region, the share 
of total personal income appears, not unexpectedly, to have ceased its rapid 
growth phase somewhat before population growth has slowed down. On the 
other hand, if convergence is upward from below the national average, as in 
the Great Lakes region around 1870 or the Plains around 1890, it seems likely 
that rapid growth in the regional share of total population has ceased some- 
what before rapid growth in the total income share. 



Income and Industrial Structure 

We turn finally to some further consideration of the possible relationships 
among demographic patterns, industrial structure, and levels of per capita 
personal incomes. The associations between urban-industrial structure and 
levels of living in 1920 were presented in Table 70. Those existing in 1950 
are shown in Table 127. 

By 1920, there was a fairly close correlation between the degrees of urbaniza- 
tion and industrialization of the labor force and levels of personal income per 
capita; a high degree of urban-industrialism meant, without exception, a high 
level of per capita income in a region. What was true of 1910-1920 had not 
been true of 1870-1880, when the highest levels of per capita personal income 
had obtained in the sparsely populated rural-mining regions of the Far West 
and Mountain states. On the other hand, the association between rural-agri- 
cultural structure and low levels of per capita income held consistently in both 
periods for regions of high and low population density. In 1950 the associations 
of 1910-1920 still hold in a general way, in spite of the continuing convergence 
of occupational structures and levels of per capita personal income; urban- 
industrial regions rank above the national averages as they did in 1910-1920, 
rural-agricultural regions rank below. The close identity of rank orders among 
the various categories, however, is no longer so sure. 

As regional levels of living and occupational structures have converged on 
national levels and structures in recent decades, deviations from the earlier 
identity of rank orders have increased somewhat. The Far West, for example, 
which again ranked first in per capita income by 1950, ranked fourth in 
urbanization, third in industrialization (manufacturing-services), and had the 



282 / Regional Economic Development: 1870-1950 



Table 127. Rank order of regions, by personal income per capita, urbanization 
of population, and labor force industry components, 1950 





Personal 


Urban pro- 


Labor force components 






% Non- 


% 




income per capita 


portion of 


% 


agric* 


Mfg.- 


Region 


(1929 dollars) 


population 


Agric. 


resource 


service 


United States 


$1,062.7 


59.0%** 


11.57 


2.03 


86.40 


Far West 


1,279.4 


62.7 (4) 


7.44 (6) 


1.87 (4) 


90.69 (3) 


Middle Atlantic 


1,255.2 


74.0 (2) 


3.09 (8) 


1.62 (5) 


95.29 (2) 


Great Lakes 


1,183.9 


65.7 (3) 


8.36 (5) 


1.01 (7) 


90.63 (4) 


New England 


1,160.4 


74.3 (1) 


3.15 (7) 


0.78 (8) 


96.07 (1) 


Mountain 


1,017.8 


51.8 (6) 


17.71 (3) 


3.98 (1) 


78.31 (6) 


Plains 


1,003.6 


49.9 (7) 


23.84 (1) 


1.03 (6) 


75.13 (7) 


Southwest 


916.6 


55.5 (5) 


16.01 (4) 


3.82 (2) 


80.17 (5) 


Southeast 


719.2 


38.0 (8) 


21.35 (2) 


3.60 (3) 


75.05 (8) 



Figures in italics are above the national average. 

* Forestry and logging, mining, and fisheries. 

**01d Census definition of urban; under new definition the proportion of urban popula- 
tion would be somewhat higher in most cases but would not affect the rank order; under 
new definition U. S. proportion is 64% in 1950. 

Rank order in parentheses. 

Source: C. F. Schwartz and R. E. Graham, Jr., Personal Income by States Since 1929, 
Table 1; see Table 8, note; 17th U. S. Census 1950, see supra, Table 4, note; and 
Appendix Tables A1-A7. 



third smallest agricultural component. The Middle Atlantic region ranked 
second in per capita income, second in urbanization and industrialization, but 
had the smallest agricultural component of any region. The Great Lakes region 
ranked third in per capita income, third in urbanization, fourth in industrializa- 
tion (actually only .06 of a percentage point below the Far West), and had 
the largest agricultural component of any of the higher income regions. Finally, 
the small and densely populated New England region in 1950 ranked fourth 
in per capita income but first in urbanization and industrialization and had 
the second smallest agricultural component. The rank order of urban-industrial 
regions in per capita income, incidentally, has not had any apparent connection 
with the size of their non-agricultural resource components (mining, forestry, 
and fisheries) , at least in the period since 1900. 

The fact that occupational structures and levels of living do not have a simple 
or too-obvious relationship in the more advanced stages of industrial growth 
is highlighted by the position among the relatively poorer regions. The Plains, 
for example, with its second lowest level of labor force industrialization, highest 
agricultural component, and second lowest level of urbanization "ought" to 
exhibit a lower level of living than the Southwest. As recently as 1950, 



Regional Adjustments within National Economy 1910-1950 (3) / 283 

however, this was not the case. It is noteworthy, nevertheless, that these poorer, 
less urbanized regions not only still have the largest agricultural labor force 
components but, with the important exception of the Plains, the largest non- 
agricultural resource components as well. The Mountain, Southwest, and 
Southeast regions, in fact, have a higher ratio of non-agricultural resource 
activity to manufactures than any of the other regions. If their further indus- 
trialization turns on the development of the earlier stages of raw materials 
processing, this kind of activity usually offers relatively low value-added per 
worker and is not likely, therefore, to cause any sudden improvement in their 
relative income positions. Now that the convergence of income levels has 
proceeded so far, further changes in relative income positions will depend on 
the kinds of activities carried on, i.e. whether low or high value-added per 
worker, and on adjustments in regional population shares. 

But here we are entering on the analysis of what lies behind the more recent 
expansion of productive activities in various regions and behind the varying 
rates of growth in income, the subjects of Parts IV and V which follow. In 
Part III, we have sought only to detail the long-run structural changes in the 
American economy as background for the economic analysis of more recent 
regional adjustments. 




Resource activities and regional 
economic growth since 1870: 

A RETROSPECT 



Since 1870 the people of the United States have experienced an unprecedented 
rise in their average level of material well-being. Over the first half of this 
century alone, while population doubled, real national output increased fivefold 
and output per capita increased two and a half times. But the actual human 
effort expended — as measured by total man-hours of labor input — increased 
by only 80%. x In the same period real personal income per capita rose by 
more than 150%, reflecting the fast-growing productive efficiency of the 
economy as a whole. Economic progress in the United States, has, of course, 
been interrupted by recurrent, and occasionally extreme, cyclical fluctuations; 
nevertheless, the long-run upward trend over the last eight decades is un- 
mistakable. 

While sharing in the nation's economic progress, the various states and 
regions have experienced diverse rates of growth, and there has been a con- 
tinuing tendency for population, productive activities, and personal income to 
be redistributed across the continental territory. The share of the nation's 
population residing west of the Mississippi, for example, almost doubled in 
the years 1870-1950. The share of total personal income payments received 
west of the Mississippi rose by about half in the years 1880-1950. Personal 
income per capita has risen in every part of the country and, more significant, 
differences in average income levels among all of the regions have been 
reduced, especially in the last quarter of a century. Certain tendencies in the 
structure of the nation's labor force have also been evident in the changing 
structure of the regional labor forces. Since 1910, for example, though not before, 
the agricultural sectors have steadily fallen and the service sectors have risen 
except in the Mountain region. And since the 1920's, even the Mountain region 
has conformed to these nationwide trends. From the standpoint of employment, 
the service sector has provided the bulk of growth. Finally, every region has 

*F. C. Mills, Productivity and Economic Progress, Occasional Paper 38 (New York: 
National Bureau of Economic Research, 1952), pp. 2-5. 

284 / 



Resource Activities and Regional Economic Growth since 1870 / 285 

become more highly urbanized than it was half a century ago and, though the 
shape of metropolitan areas has changed since 1920, there seems little likelihood 
that the urban tide will soon ebb. 

The fact that rates of urbanization or of growth in per capita income are 
usually higher in regions that were still "undeveloped" in the late nineteenth 
century means that the economic growth of the nation has reduced disparities 
in levels of living among the regions. Judged by the convergence of personal 
income per capita, however, the movements of productive resources to and 
from the more affluent regions have not yet left any of the less affluent regions 
at a permanent disadvantage. There have been currents of demand and counter- 
flows of capital, for example, which have served to increase production in 
many of the less developed parts and have helped raise average levels of well- 
being. All regions, therefore, contribute something of importance to the 
continental commonwealth and in turn share in the general gain. 

If this convergence of demographic, labor force, and income structures among 
the regions represents the principal finding of the historical section of this 



o ooo o 
r- a) r-n in 

CO CO ™ O) 01 



/Qua 

\ooooo 

J* CO S 2 "> « 




Figure 42. The Changing Regional Structure of the Labor Force, 1870-1950. 
Source: Appendix Table A-l. 



study, it is of almost equal significance that regional differentials have not 
been entirely eliminated (see Figure 42). Regional differentials (and within 
regions, state differentials) in the structures of productive activities, in per 
capita output and income, and in salary-wage rates remain, in spite of the 
over-all tendency to convergence. 

Were it not for the relatively high average levels of personal income that 



286 / Regional Economic Development: 1870-1950 

prevailed in the Far West and Mountain regions during the closing decades 
of the last century, it would appear that the more urban-industrialized parts 
of the country have been "better-off" than the rural-agricultural regions over 
the entire period since 1870. Indeed, they have been "better off," for the Far 
West and Mountain regions at that time were not so much an exception to the 
rule regarding differentials between urban and rural regions as a distinctive type 
of region not covered by the urban-rural rubric. Thus the agricultural com- 
ponents of the labor forces in the two regions were far below the agricultural 
component of the national labor force; they were not rural-agricultural regions 
of the same type as the Southeast, the Plains, or even the Great Lakes. Nor 
were they, on the other hand, manufacturing regions such as the New England 
or Middle Atlantic regions. Rather were they remote, sparsely populated areas 
with large proportions of their relatively small labor forces engaged in service 
occupations (especially transportation) and resource activities other than agri- 
culture (mostly mining and lumbering). Already in 1870 the Far West was the 
third most urbanized region in the country; and some of the Mountain states, 
Colorado, for example, were urbanized to a surprising degree. 

Since 1900 the substantially higher levels of per capita personal income 
achieved in the more urban-industrial regions have tended to raise the average 
level for the nation. The differential experience of the Far West and Mountain 
regions during the decades 1900-1930 underlines this fact. After 1920 the 
Mountain region fell below the country-wide average, this decline coinciding 
with the years in which its agricultural labor force component reached an all- 
time high. Though the Mountain region's per capita income levels have been 
converging upwards again since 1940, its fortunes have more closely paralleled 
those of the agricultural Plains rather than those of the more diversified Far 
West. The rising levels of per capita personal income in all of the agricultural 
regions in recent decades may stem as much from the fact that they have become 
more highly specialized parts of the greater industrial economy of the nation 
as from their higher local concentrations of urban-industrial population. 
Broadly speaking, urban-industrialization appears to have been the dynamic 
ecological pattern assumed by growing populations, regionally and nationally, 
as they have organized to control, utilize, and enjoy a greater volume and 
variety of material goods and services. 2 

Certainly, the economic and social transformation implicit in the redistribu- 
tion and restructuring of the labor force and productive activities among the 
regions after 1870 was a product of rapid industrialization, a process which 
had already gathered momentum in the Northeastern part of the country during 
the second quarter of the nineteenth century. Continued industrialization 
eventually required the mobilization of resources on a continental scale, thereby 
heightening territorial specialization and regional interdependence across the 
entire nation. Under the technological and organizational conditions of manu- 
facture prevailing at that time, urban concentration was, so to speak, the 

2 In this regard, see the suggestive hypotheses of J. P. Gibbs and W. T. Martin, "Urban- 
ization and Natural Resources: A Study in Organizational Ecology," American Sociological 
Review, Vol. 23, (June 1958), pp. 266-77. 



Resource Activities and Regional Economic Growth since 1870 / 287 

demographic corollary of rapid economic growth. Apart from the increase in 
population itself, the crucial variables in the industrial process were technology 
and organization. Technology, of course, means the practical arts, the physical 
and intellectual capital accumulated by a resourceful population in order to 
achieve its higher levels of living: the technologies of transportation, of rapid 
communications, of manufactures, and, not least, of resource extraction, notably 
the technologies of agriculture and mineral development. The distinctive role 
of organization, however, should not be underestimated even though it defies 
close measurement. It signifies the complex institutional arrangements, the 
modes of conducting business, the forms of community structure that arise 
among populations differentiated, functionally and spatially, in the performance 
of their more specialized and routinized tasks. 3 Organizational functions (for 
example, the proliferation of independent "middlemen" and formalized bu- 
reaucracy) have become major contributors to the growth of service activities; 
they systematize the wider market and give direction and coherence to the ex- 
panding social economy. 

In addition to population, technology, and organization, a fourth variable 
is required to round out the analytic framework of regional economic growth, 
namely, environment. Natural resources are a special feature of the environ- 
ment; but, in the broadest sense, the term refers to the total milieu in which 
a population subsists — all the external forces affecting life, including habitat 
or dwelling-space and other populations. We have been primarily concerned 
with the resource potentials of different regional environments but have argued 
that natural resources do not acquire significance for economic growth until 
they are confronted by a highly motivated population equipped with technology 
and organization to develop them. The interdependence among the four 
variables is implicit in our use of the metaphor "the play of technology on 
resources." This interplay accounts for both the wider and the more efficient 
use of resources. Americans organized and applied technology in myriad ways 
to make the resources of nature useable and available. Whether or not such 
resources were used at any given time was largely determined, of course, by 
conditions of the market. Thus the longrun impetus to the development of the 
agricultural, mineral, and forest potentials of the greater Southeast and West 
was the expansion of the urban-industrial economies of Western Europe and 
the Northeastern United States during the nineteenth century; capital and 
population were exported from both these "advanced" regions in order to 
render the hitherto untapped resources of undeveloped regions accessible and 
available to world markets. 

The territorial expansion of the American economy, therefore, cannot be 
understood without recognizing the interplay of technology and environment 
in meeting the growing demands of urban-industrial populations. The growth 
of that economy, moreover, cannot be explained historically without considering 

3 K. E. Boulding, The Organizational Revolution (New York: Harper and Bros., 1953), 
pp. xvi-xxiv, 202-221. Also, T. W. Schultz, The Economic Organization of Agriculture (New- 
York, McGraw-Hill, 1953), which places emphasis on the relation between urban centers 
and economic development. 



288 / Regional Economic Development: 1870-1950 



Regional Proportions of total U.S. Primary Employment 



1870 




F,W. 2% 

NX 5% 

M.A. 15% 



6,684,219 




13,596,815 



1950 

F.W. 7%-y rN.I.2% 
S.W.10% 




8,184,509 



Regional Proportions of Respective industries 



Agriculture 

;0 20 30 




forestry 



m. 



W V///////A 



Fishing 



Northeast 
Southeast 



West 



fficufture 

10 20 30 



m\ 






Northeast 












West 







Mining 




W&fl 




MAf : x : ' ; " : : : : : : : :: : :v: : : : : : : : : : : : : : : : 




w 


„ J^ Southeast 

Wesf 



Forestry 




wimZZZMk 



Fishing 




Northeast 



Norfheasf 

Southeast 



Agriculture 

J2 £2 3P- 




Mining 



!!£= 



\ Northeast 

l .Southeast.. 




Figure 43. National and Regional Employment in the Primary Industries, 
1870, 1910, and 1950. 

Source : Appendix Tables A2-A5. 



the geographical extension of its effective boundaries and the incorporation of 
"new" regions into the system of national and international markets. These 
two features of the period 1870-1910 were, of course, a continuation of a 
process dating back to the original settlement and development of the maritime 
sections of North America during the seventeenth and eighteenth centuries. 
The "triangle trade" of the colonial period, the swelling currents of exchange 
between the grain producing "Northwest," the cotton-producing "South," and 
the industrializing "Northeast" during the early nineteenth century, climaxed 



Regional Activities and Regional Economic Growth since 1870 / 289 




Figure 44. Regional Proportions of Value of Primary Resources Extracted. 
1870-1950. 

Source: Appendix Table B-l. 



in a system of multi-lateral regional interdependence by the close of the century. 
The continuing extension of accessibility has been a critical factor in the growth 
of the economy. The exploitation of the contiguous land mass of the interior 
was always dependent on great improvements in transportation and communi- 
cations. Before 1840 the economy was for all practical purposes coterminous 
with the capacities of coastal and inland waterways. Prior to the completion 
of an integrated railroad network during the last quarter of the nineteenth 
century, the full resource potential of expanses situated between the Mississippi 
system and the Pacific coastal fringe could not be made available for economic 
growth. No doubt, the possibility of occupying and developing the vast Western 
territory provided a powerful incentive to individual and corporate enterprise 
at home and abroad. Eventually the greater West was to augment the quantity 
and variety of resource availabilities; nevertheless, failing the requisite tech- 
nique and organization to integrate the "new" resource regions into the older 
Atlantic economy, demand for such resources would have been ineffective and 
their development retarded. 

Throughout the period under review, there has been a substantial redistribu- 
tion of resource activities across the nation (Figures 43 and 44). This is only 
to be expected, since many older sources of materials have been depleted, new 
ones have been developed, and the whole resource economy has been trans- 
formed by technological advances which have given rise to new resource uses 
or have reduced the volume of resource inputs in a given manufactured output. 
The redistribution of activities was somewhat greater in the first half of the 



290 / Regional Economic Development: 1870-1950 

period. The consumption of many raw materials per unit of gross national 
product has generally declined during the present century, 4 and the location of 
manufactures has in general become less oriented towards raw materials and 
more towards markets. All but three of the eight regions — the Plains, the 
Mountain, and the Far West — have smaller shares of the nation's total resource 
employment than they had before 1910. Only two regions, moreover, show 
sizable increases in the proportions of the labor forces employed in major 
resource activities: the Southeast shows an increase in the relative size of its 
labor force in both forestry and logging and mineral extraction, the Southwest 
has a larger proportion in mineral extraction alone (see Figure 43). 

Compared with the period before 1910, the redistribution of agricultural 
activities has been slight: a relatively minor decrease in the Southeast and a 
somewhat larger concentration in the greater West outside of the Southwest. 
The major belts of farm staples are unchanged except at their perimeters; 
high-value crops have continued to locate close to the major urban markets 
or in a few select localities with special climatic advantages; the more spec- 
tacular increases in farm productivity have been largely confined to these same 
localities. 5 Forestry and logging activities have declined relatively in the 
Northeast, outside of New England, and have remained relatively stationary 
in the greater West; the share of the Southeast meanwhile has risen by about 
the same proportion as that of the Northeast has declined. Mineral activities 
reveal the greatest redistribution since 1910. The Northeast has declined 
relatively by almost 40%, the greater West has risen by about 17%, mostly 
in the Southwest, while the Southeast has more than doubled its share of 
mineral activities. 

Most of the redistribution in mineral activity can be attributed to the change 
in the relative position of coal and oil products in the nation's fuel economy, to 
the growth of the Southeast and Southwest, and to the relative decline of coal 
mining in the Middle Atlantic region. The principal sources of petroleum and 
natural gas since World War I have been located outside the Northeastern 
manufacturing belt, notably in the Southwest, the Far West, and Gulf areas 
of the Southeast. Coal, on the other hand, is still obtained in large quantities in 
the Middle Atlantic and Great Lakes regions and in adjacent parts of the South- 
east, chiefly West Virginia. Coal and coke remain the major sources of heat- 
energy in two of the nation's most industrialized regions, hence the most power- 
ful stimulus to petroleum consumption so far has not been industrial manufac- 
tures but the automobile. 

To the extent that resource activities have contributed to the convergence of 
per capita personal income among the regions in recent decades, the growth 
of petroleum activities and declines in staple agriculture have probably made 
the greatest contributions. 

4 According to indexes of consumption of select materials per unit of G.N.P. made by 
E. Vera Eliasburg, Resources for the Future, Inc., only oil and gas and non-ferrous metals 
among major resource categories show any sizable increase since 1900. 

5 U. S. Census of Agriculture, 1950, Vol. V, Special Reports, Pt. 6, Agriculture 1950 — A 
Graphic Summary (Washington: Government Printing Office, 1952), 69-102. 



Regional Activities and Regional Economic Growth since 1870 / 291 

Regional resource activities within the growing national economy since 1870 
may be summarized as follows. The initial contributions of the relatively un- 
developed Southeastern and Western regions (mostly agricultural staples and 
metallic ores) helped sustain growth already achieved in the more advanced 
urban-industrial regions on the basis of their indigenous endowments of min- 
eral, agricultural, and forest resources. By the close of the last century, resource 
depletion in the more developed Northeast, the progress of technology and, of 
course, the continuing growth and migration of population required that the 
material base of the national economy be extended over a wider area. By that 
time the area of economy more or less coincided with the political boundaries 
of the nation and henceforth growth has turned on the organization of produc- 
tion and distribution on a truly national scale; that is, on the further refine- 
ment of the system of interdependence among all the regions. Though there are 
still differential advantages for production and consumption to be had within 
the national framework, the progressive equalization of per capita income 
among the regions at higher levels, not excluding the Southeast, is proof that 
the national economy as a whole is potentially greater than any of its parts, 
provided the parts are differentiated along lines of comparative advantage. All 
regions have eventually come to share in the larger abundance. 

If the play of technology on resources has been a primary factor in the 
growth and redistribution of the nation's wealth, a necessary condition has 
been the opportunity for labor and capital to flow freely throughout the com- 
monwealth, seeking their most profitable employment in resource development, 
manufacture, or servicing, as the case may be. In the course of this general 
amelioration some regions, New England for example, have made painful ad- 
justments; wide areas of the country have been ruthlessly and in some cases, 
perhaps, needlessly denuded of their natural endowment. Coal mines have been 
exhausted, rich ores depleted, forests cut down, and the soil impoverished. Yet, 
on balance, the American economy, if sometimes terribly wasteful, is not simply 
a predatory system. In one sense, resources unused are no better than re- 
sources abused. Public and private interests have become aware that, having 
used up the riches of one area, they can no longer expect to return that area 
to nature and move on to "new frontiers" elsewhere. The frontiers of tomor- 
row must be found in their own backyard. 

The same thrusts of technology and the same, if more civilized, instincts 
which urged population out across the continent in search of higher returns 
have also reduced the quantities of resource inputs required for a given pro- 
duction: new and more economical methods of extraction, precision control 
over the processing of materials, and a nationwide organization of production 
and distribution have continually yielded higher returns to effort. Public and 
private programs for conservation over the last half-century have, in the long 
run, contributed to the same end. In recent years, however, the American 
economy has drawn increasingly on the natural resources of countries which 
do not immediately share in the full prosperity of the American system. It 
cannot be known how critical this dependence is likely to become. The nation's 
own regional experience, however, indicates that, in spite of technological 



292 / Regional Economic Development: 1870-1950 

economies at home, the newer interdependence reaching across national bound- 
aries is likely to be more rather than less important in the years ahead. It is 
in the resource sector above all that the need for interdependence is likely to 
be felt by the United States. Whether Americans can apply the institutional 
lessons of their own good fortune in their economic relations with others is a 
question that in all probability will be answered not too far in the future. 



Part 
IY The Regional Distribution of 

Economic Activities in the 

United States, 1939-1954 



IQ / An approach to the study 

of changes in the regional distribution 
of economic activities 



The broad regional changes in economic activity between 1870 and 1950 that 
we have described bear out the observation with which our study opened — that, 
although the United States has grown steadily in population and per capita in- 
come, the various regions within the national economy have not all shared 
equally in this growth. Some have shown a rate of development exceeding the 
average for the nation as a whole; others have fallen below the national stand- 
ards. 

To arrive at an understanding of this phenomenon, we shall now inquire 
more deeply into the principles governing regional distribution of economic 
activity. We are primarily concerned at this point with the regional distribu- 
tion of the volume of economic activity (rather than interregional differences 
in welfare). The best over-all view of this distribution is provided by employ- 
ment data, for they enable us to study in detail what has been taking place 
in specific sectors of economic activity. Most of our attention here is addressed, 
therefore, to analyzing the meaning of employment structure and employment 
shifts in terms of differential regional growth. In our examination of these 
data we limit our inquiry to the years 1939-1954, both because this period 
is more specifically relevant to an understanding of present regional positions 
and because it permits us to use more detailed techniques and reach deeper 
levels of analysis of the phenomenon with which we are concerned. 1 

a Although employment does not reflect all of the factors determining the comparative 
economic importance of industries, it provides the only uniform basis for measuring and 
comparing the distribution of all types of economic activities. Furthermore, in some ways 
employment is a better measure of relative activity than either "value of products" or "value 
added," which are available for manufacturing only. The former includes the value of raw 
materials; and for both, comparisons between different periods would be distorted if changes 
in the price level took place. (It should be noted, however, that this could be corrected 
for.) While employment alone is far from a perfect measure of relative changes in volume 
of activities, it would seem to be the best measure, and employment changes are also signifi- 
cant in and of themselves. 

/ 295 



296 / Regional Distribution of Economic Activities, 1939-1954 

Employment changes are particularly useful as an indication of changes in 
the volume of economic activities because of the close tie between employment 
and population. The rank correlation between total population distribution by 
states and total employment by states is .991; clearly the association is ex- 




Absolute upward shift 
1 Relative downward shift 



Figure 45. Net Shift in Total Employment, 1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 18.5. Total net shift as % of 1939 employment: 7.1. 
Source: Appendix Table H. 



tremely close. Figure 45 shows the net shifts in total employment by states 
from 1939 to 1954. When this is compared with Figure 10, showing the net 
population shifts for the same period, the similarity of configuration is evi- 
dent. In all but five states, 2 the relative shifts in population and in employment 
take the same direction. 

2 Kansas, Louisiana, Virginia. Illinois, and Tennessee. However, it should be noted that 
while the correlation between population and employment is very high, individual variations 
can be significant. There are thirteen states where employment share deviates from popula- 
tion share by as much as 10% or more. There are many reasons why this is so. (1) In 
some cases there is significant data distortion. Employment data are subject to some bias 
because of interregional variation in business cycle performance. There is also reporting 
bias. In heavily agricultural states, employment share is inclined to be overstated because 
of the tendency to report as employed farm workers who are, in the words of Rudyard 
Kipling, "scrutiating idle." (2) In other cases the discrepancy between employment share 
and population share is a result of the influence of amenities. Many households can locate 
independent of the requirements of economic opportunity. This appears to be clearly the 
case for Florida and Arizona, for example. In spite of the fact that these states had low 
shares of income and employment relative to population, they continued during 1939-1954 



An Approach to the Study of Changes in Distribution / 297 



Changes in Regional Activity Revealed by Total Employment 

We have already presented in Figure 45 a summary picture of the net shift 
in total employment among the states between 1939 and 1954. 3 Table 128 
shows the relative importance of the shifts over these years in the ten broad 
industry sectors or 1 -digit components of total employment. The shifts amounted 
to a significant proportion of the absolute changes in employment in each 
of the major employment sectors, averaging 18.5% for all, (Col. 8). However, 
only in three sectors — agriculture, mining, and finance-insurance-real estate — 
were the employment changes among some of the states greatly removed from 
the national average for the sector. And only in mining was the change in 
position among the states really great compared to the national average ab- 
solute change. In terms of the size of the net shifts as a percentage of the 
number of persons employed in the category in 1939, construction is in the 
"large shift" class also (Col. 9). 

We shall now turn to the employment shifts shown in Figure 45 to see 
whether some useful generalizations can be derived from them. But this exam- 
ination will be much more revealing if we recognize the distinction that we 
have previously made between the "differential" shifts and the "proportion- 
ality" shifts that together make up the total employment shifts. By way of re- 

to augment their population shares at a more rapid rate than their shares of employment. 
(3) In still other cases the discrepancy may be a symptom of a lag in adjustment to 
economic opportunity. In any case, if the reader wishes at some point to apply the demon- 
strations that follow concerning structure and change in employment to an understanding 
of the distribution and shifts in population, he will do well to recognize the existence of 
and reasons for these discrepancies. 

throughout Part IV we shall place considerable emphasis upon shifts in employment 
patterns as evidence of the changes taking place in the regional structure of the American 
economy. Our repeated use of this device, however, should not lead the reader to overlook 
two very important facts. First, when we concentrate upon the "shift" pattern, we are 
focusing attention upon the marginal increment of change in a regional economy. Often 
this is only a small part of the total absolute increment demonstrated by the dimension 
being examined. For example: In the case of employment over the period 1939-54, the 
total net shift in position among the states was slightly less than 3,000,000 wage jobs. This 
was only 18% of the total increase in employment during this period, and only 7% of the 
total 1939 employment. These are significant shifts, to be sure, and well worth examining. 
But the use of these data has consequences that need to be kept in mind. 

(a) Constant attention to these incremental changes may give an exaggerated notion of 
the quantitative significance of regional shifts. The truth is that the American economy 
demonstrates enormous stability in the face of change. At the same rate it would take 200 
years for employment shifts across state boundaries to equal total 1939 employment. At the 
present time, two-thirds of the nation's labor force is employed in the states east of the 
Mississippi that make up only one-third of the total land area. The states in the manu- 
facturing belt, with less than 15% of the nation's land area, contain roughly 45% of its 
population and employment. The shifts we describe here are not apt to change these 
proportions significantly within the short-run future. 

(b) Focusing attention upon net shifts magnifies the possibility of error resulting from 
imperfect data. It is possible, therefore, that some of the shifts we describe have a 
spurious origin. This is more likely when a state can claim only a very small portion of 



298 / Regional Distribution of Economic Activities, 1939-1954 

view, the differential shift is the shift which a region shows relative to the na- 
tional average in employment in the broad industry components (the so-called 
1-digit categories) of total employment. 4 In other words, it is the intra-industry 
effect among regions. Thus the principal "growth" regions may show their net 
upward shifts because, for a variety of reasons, they are outstripping the rest 
of the nation in their rate of increase in the components of total employment. 
The proportionality shift, on the other hand, is the effect of a region's indus- 
trial composition in terms of the employment sectors that, for the nation as a 
whole, have expanded more rapidly or less rapidly than others. The "growth" 
region then may be one that has a proportionately larger share of the rapidly 
expanding sectors of employment. The total net shift in employment shown in 
Figure 45 is the sum of these two types of shifts. 

There are statistical counterparts to these concepts. If we compute the net 
shifts for each of the ten 1-digit components of total employment, we find that 
the sum of the differential net shifts displayed by each component for a single 
state rarely exhausts the total net shift in employment. The balance, then, is 
the proportionality net shift. In Figures 46 and 47 we divide the total net shift 
presented in Figure 45 into its differential and proportionality components. 5 
It works out in this case that about half of the total shift is generated by the 
proportionality effect and about half by the differential effect. 

Figures 46 and 47 reveal how widely these two types of net shifts differ in 
their regional effect. About 92% of the total proportionality net upward shift 
has occurred in the Manufacturing Belt states. Outside this area, only five 
states — Maine, New Hampshire, Florida, California, and Washington — ex- 
hibit net upward proportionality shifts; and of these, only California can claim 
a sizeable net gain from the proportionality effect. In contrast only Michigan, 
Indiana, and Ohio in the Manufacturing Belt show net growth from the differ- 
ential effect. The net differential upward shift is concentrated in the part of 
the nation west of the Mississippi, except for a few of the Plains and northern 

the total net shift. Generally, however, this type of error can be avoided by careful atten- 
tion to other data and types of knowledge. 

It is also well to note with regard to the use of the shift analysis that, as the gains or 
losses in employment shares are based upon relative rates of change, it is quite possible for 
a state with a "net downward shift" (i.e., a below-average rate of increase) to have a 
larger incremental increase in employment than one experiencing a "net upward shift." This 
is a function of the differences in the absolute size of the economic base. The focus on rela- 
tive changes as against a national standard is appropriate for the principal questions being 
examined here. There are, however, a number of questions for which a knowledge of 
absolute elements might be more relevant. 

4 The breakdown employed here covers: agriculture; mining; construction; manufacturing; 
transportation and public utilities; wholesale trade; retail trade; finance, insurance, and 
real estate; service and miscellaneous; and government. 

B The reader should be aware that the statistical dimensions of the proportionality and 
differential shifts (or total net shift for that matter) are not independent of the arbitrary 
decisions concerning the way and the frequency with which we slice a given employment 
universe. If we select a different classification of ten components we will get different shift 
dimensions. If we increase the number of components to 20 we will get different shift 
dimensions. We point this out to emphasize that such measures can be interpreted only in 
terms of the specific classifications upon which they are based. 



An Approach to the Study of Changes in Distribution / 299 




f^Mj Upward shift 
I | Downward shift 



Figure 46. Differential Net Shift in Total Employment, 1939-54. 




Upward shift 
I I Downward shift 



Figure 47. Proportionality Net Shift in Total Employment, 1939-54. 
Source: Table 157. 



300 / Regional Distribution of Economic Activities, 1939-1954 



V 

X 



to 




u 




a 




2 








tn 




O 




Ch 


. 


B 


LO 




C\ 


CO 

> 


r— 1 


*4J 


On 


_5 


(TO 




Cs 




C/T 




5h 


cc 


O 


CS| 


O 


l-H. 


CO 




t« 


45 




"-— 


■4-) 


-o 


s 


e 


03 


^ 


S 



On r-J O ■ — i 

^ © to O 



~3 a °1 



43 




T^ 




^ C 


4 _, 


LO 




t— _ 


<-M 


0\ 




— ' « 


IS 


ro 






0!) 


ON 




O 








H 






<-M 


~v a 


03 


LO 


° ^H 


VO 1 — 1 




^ 


^~ 


"# 


U 


o 


o 

CO O 



3 . LO 

^ "o J« S5 

rQ ° ON 

<! rH 



w 



^ 



4 _ ) 


tF 


PI 


LO 


CD 


on 


E 


1 — 1 






>> 




_o 




"a, 




£ 





^ 



a 

^1 8 

S 

03 



r-j 


H . O N «3 


i> CM c^ 


r> lo # <~i 


!>* 


m o m d 

NtNH 


t~~* r-i t>: 

T-4 


V© !>■" <tf 

rH rH 


LO 


t> O CM CO 


cq "* cm 


N O <* 


CO 


^ VO o6 o 
CO CM S rH 
CM 


o< cd vd 


© r-i C3 
CO rH CM 



VO 
CO 


CO o o CO 
O On ON ON 
Oy CO t^ CO 


O CO CO 

co r^ csi 

rH rH CO, 


rH O ^ 

$ 3 £. 


vo" 

ON 
Cm" 

+1 


csf lo" r-T vo 

N ^O \C CO 
LO i— 1 CM ON 


co" On' CO 

cm r~- r- 

(N H ro 


n oC w* 

CM t? VO 
CM CM LO 


+1 +1 +| +1 


+1 +1 +1 


+1 +1 +1 



I I 



i> cq o o 


N CO h; 


cm vq c-q 


-* CO CM CO 


CO 6 N 


LO* CM ON 


r-H CM VO 


CO MO tJF 


LO ^ NO 



I I 



1 

CO 


,652,909 

73,172 

,435,600 

,060,394 


,127,600 
975,660 
,299,525 


753,700 
,090,300 
,759,500 


lo" 


<-t rH VO 


rH CM 


CM CM 



*£ 


On cq vq r~- 


rH VO VO 


t-~ vq on 


© 


NO r-i -Hh" r-" 


t> tF CM 


CO* ON t-A 


O 


y-t CM 


r—t 


i—l 



© 


VO CO © t^ 


O t^ rH 


o o o 
o o o 
•># o o 


CO 


rH CO O rH 


O O LO 


CO 


cq CM vo co 


VO O CO 


On" 


N ^~ N fO 


ON rH rH 


rH 7 g $ 


CO 


On lo ^h CO 


CO CO CM 


VO 


in N VC; MD 


O LO rH 


rH ^ t> 


VO 


On" CM lo" 


<*" csf r-" 


C<f lo" vo" 


LO 


1—1 







LO O ON LO 



rH On CO 
t^ CO* rH 



CO 


CM O 


O CO 


O l> VO 

O xrjl O 


CO 


LO 1-^i 


O CM 


CM 


CM «* 


O ON 


O CO 00 










8 


o r- 


r- cm 


CM LO rH 


LO CM 


r- cm 


rH O CM 


On 


CM CO 


r-^ VO 


On VO CO 










©" 




rH ON* 


C<T i-h" <* 


"«* 


rH 







CO CM CO 

CO CO* On 



s s ° 

p p ° 

t^ l> r-^ 

r?" oC t~J~ 

bS £2 °° 

CO &\ ON 

rH CO" CO" 



S 



ro S 13 



03 

S tie 



•2 2 5^^ 



C ^H 



03 g 



0^ 



O ^ rJ S S 2 '« ^ 03" S c« 

ft g Cil^ »H^S C13." 

a .— "-i c» a <-< i-i ^ 3 e r* T 1 



B3 o * S 
<J S U S H 



c a a o «j 
pH^a 



tr 03 .3 

r^ tf £ 



03 

D o 
V2 O 



03 

H 

a 

03 
ft 



An Approach to the Study of Changes in Distribution / 301 

border states. The upper Southeastern states, however, share this effect. Cali- 
fornia, Texas, and Florida led the nation in differential rates of growth during 
the period under review. 

Table 129 shows how the total shift for each state is divided between the 
proportionality and differential effect. There is considerable variety of per- 
formance. For example, Florida, the Far Western states, the southern Mountain 
states (Colorado and Utah), and Arizona demonstrate net upward shifts almost 
entirely because their important employment sectors are growing at a rate that 
exceeds the national average for these sectors. Maine and New Hampshire 
exemplify the reverse of this situation; they show net downward shifts because 
their important employment sectors have declined or have failed to keep pace 
with the rate of increase for these sectors in the nation as a whole. 

In contrast, specialization in employment sectors that were growing at a 
faster rate than the average for all industries largely accounts for Maryland's 
net upward shift in total employment. And specialization in employment sectors 
that were declining nationally or growing at a below-average rate for the nation 
as a whole is the major factor in the net downward shifts shown by Minnesota, 
Iowa, North and South Dakota in the Plains region, by Wisconsin in the Great 
Lakes region, and by Kentucky, Arkansas, Mississippi, Alabama, and South 
Carolina in the Southeast. 

In all the other states the total employment shift significantly reflects both 
the proportionality and the differential effect. In some cases (Michigan, for 



Table 129. Total, differential, and proportionality net shifts in employment, 
by state, 1939-1954 



State 




Absolute shift 




% of 


state total 




(1) 


(2) 


(3) 
Propor- 


(4) 


(5) 
Propor- 






Differ- 


tional- 


Differ- 


tional- 




Total 


ential 


ity 


ential 


ity 


New England 












Maine 


- 34,687 


- 40,091 


5,404 


-115.6 


15.6 


New Hampshire 


- 36,155 


- 55,243 


19,093 


-152.8 


52.8 


Vermont 


- 24,371 


- 18,778 


- 5,593 


- 77.1 


- 22.9 


Massachusetts 


-119,129 


-368,577 


249,448 


-309.4 


209.4 


Rhode Island 


- 50,832 


-101,772 


50,940 


-200.2 


100.2 


Connecticut 


35,728 


- 70,713 


106,441 


- 197.9 


297.9 


Middle Atlantic 












New York 


- 79,626 


-777,501 


697,875 


-976.4 


876.4 


New Jersey 


123,322 


-100,699 


224,021 


- 81.7 


181.7 


Pennsylvania 


-276,399 


-545,910 


269,511 


-197.5 


97.5 


Delaware 


14,919 


5,943 


8,976 


39.8 


60.2 


Maryland & D.C. 


105,893 


- 49,567 


155,460 


- 46.8 


146.8 



302 / Regional Distribution of Economic Activities, 1939-1954 
Table 129— continued 



State 


(1) 


(2) 


(3) 


(4) 


(5) 


Great Lakes 












Ohio 


304,790 


128,436 


176,354 


42.1 


57.9 


Indiana 


63,052 


35,339 


27,713 


56.0 


44.0 


Illinois 


498 


—249,193 


249,691 


-50,039 


50,139 


Michigan 


227,725 


98,019 


129,706 


43.0 


57.0 


Wisconsin 


- 52,577 


- 14,966 


- 37,611 


- 28.5 


- 71.5 


Southeast 












Virginia 


- 13,591 


36,141 


- 49,732 


265.9 


-365.9 


W. Virginia 


-159,903 


— 88,733 


- 71,170 


- 55.5 


- 44.5 


North Carolina 


-103,806 


41,250 


-145,056 


39.7 


-139.7 


South Carolina 


-211,282 


- 78,006 


-133,276 


- 36.9 


- 63.1 


Georgia 


-273,931 


-115,711 


-158,220 


- 42.2 


- 57.8 


Florida 


250,586 


230,706 


19,880 


92.1 


7.9 


Kentucky- 


-118,309 


23,258 


-141,567 


19.7 


-119.7 


Tennessee 


14,034 


129,041 


-115,007 


919.5 


-819.5 


Alabama 


-239,053 


- 62,175 


-176,878 


- 26.0 


- 74.0 


Mississippi 


-299,100 


- 36,903 


-262,197 


- 12.3 


- 87.7 


Arkansas 


-211,474 


- 13,987 


-197,487 


— 6.6 


— 93.4 


Plains 












Minnesota 


- 80,226 


- 1,054 


- 79,172 


- 1.3 


- 98.7 


Iowa 


-117,559 


- 18,593 


- 98,966 


- 15.8 


- 84.2 


Missouri 


- 97,364 


- 53,463 


- 43,901 


- 55.5 


- 44.5 


North Dakota 


- 71,289 


- 2,393 


- 68,896 


- 3.4 


- 96.6 


South Dakota 


- 39,548 


4,520 


- 44,068 


11.4 


-111.4 


Nebraska 


— 30,847 


21,314 


- 52,161 


69.1 


-169.1 


Kansas 


28,708 


92,037 


- 63,329 


320.6 


-220.6 


Southwest 












Oklahoma 


- 79,746 


19,168 


- 98,914 


24.0 


-124.0 


Texas 


112,286 


408,869 


-296,583 


364.1 


-264.1 


New Mexico 


36,201 


59,403 


- 23,202 


164.1 


- 64.1 


Arizona 


68,058 


81,238 


- 13,180 


119.4 


- 19.4 


Louisiana 


- 36,947 


63,854 


-100,801 


172.8 


-272.8 


Mountain 












Montana 


- 45,079 


- 17,927 


- 27,152 


- 39.8 


- 60.2 


Idaho 


- 8,246 


20,045 


- 28,291 


243.1 


-343.1 


Wyoming 


- 4,986 


3,892 


- 8,878 


78.1 


-178.1 


Colorado 


39,695 


51,779 


- 12,084 


130.5 


- 30.5 


Utah 


43,631 


48,968 


- 5,337 


112.2 


- 12.2 


Far West 












Washington 


68,812 


60,810 


- 8,002 


88.4 


11.6 


Oregon 


48,839 


58,317 


- 9,478 


197.5 


- 97.5 


California 


1,303,805 


1,133,188 


170,617 


86.9 


13.1 


Nevada 


25,478 


26,290 


- 812 


103.2 


— 3.2 



Source: Table 157. 



An Approach to the Study of Changes in Distribution / 303 

example) both effects contribute to a net upward shift. In other cases (West 
Virginia, for example) both contribute to a net downward shift. In still other 
cases (Illinois, for example) the proportionality and differential effects tend to 
offset each other and the total net shift depends upon which effect is the 
stronger. 



Interpretation of Observed Regional Behavior 

Table 130 compares the percentage change in employment in each of the ten 
broad industry sectors with the rate of change in the nation's total employment 
between 1939 and 1954. Mining and agriculture alone fall below the national 
average for these years. In each of the remaining eight sectors the rate of 
increase exceeds the average rate for the nation's employment as a whole. It 
should follow that the states which specialize in mining and agriculture would 
show net downward proportionality shifts. 

This is precisely the case. Table 131 shows the distribution of each state's 
total employment in 1939 between the industry sectors with an above-average 
rate of growth and those with a below-average rate. Without exception, the 
states that have a large proportion of their total employment in the slow-growth 
sectors — agriculture and mining — correspond exactly to those that are shown 
to have net downward proportionality shifts in Figure 47. Mississippi, which 
has a large net downward proportionality shift, devoted nearly 75% of its 
employment to agriculture and mining in 1939, as compared with 29% for 
the nation as a whole. In West Virginia, which also shows a net downward 
proportionality shift, almost 21% of the total employment was in mining as 
compared with a national average of only 2%. In contrast, New York, which 
shows a large net upward proportionality shift, devoted 93% of its total 
employment in 1939 to the rapid-growth sectors of employment, as compared 



Table 130. Percentage change in employment in the United States, by indus- 
try, 1939-1954 



Construction + 122.0% 

Government + 69.2 

Manufacturing + 63.0 

Service and Miscellaneous + 62.6 

Wholesale Trade + 60.8 

Finance, Insurance and Real Estate + 55.2 

Retail Trade + 47.7 

Transportation and Public Utilities + 38.7 

Total All Employment + 38.5 

Mining — 8.8 

Agriculture — 14.7 

Source: Appendix Table G. 



304 / Regional Distribution of Economic Activities, 1939-1954 

Table 131. Relative employment in rapid-growth and slow-growth industries * 
in the United States, by state, 1939 



State 



Slow-growth 



Rapid-growth 



United States 

New England 

Maine 

New Hampshire 

Vermont 

Massachusetts 

Rhode Island 

Connecticut 

Middle Atlantic 
New York 
New Jersey 
Pennsylvania 
Delaware 
Maryland and D. C. 

Great Lakes 

Ohio 

Indiana 

Illinois 

Michigan 

Wisconsin 

Plains 
Minnesota 
Iowa 
Missouri 
North Dakota 
South Dakota 
Nebraska 
Kansas 

Southeast 

Virginia 

W. Virginia 

North Carolina 

South Carolina 

Georgia 

Florida 

Kentucky 

Tennessee 

Alabama 

Mississippi 

Arkansas 

Southwest 

Oklahoma 

Texas 

New Mexico 

Arizona 

Louisiana 



7o 
29,5 

27.4 

15.9 

36.6 

5.2 

2.8 

7.6 

7.3 

6.0 

17.5 

17.7 

9.8 

18.1 
25.8 
15.0 
19.5 
34.7 

40.9 
46.4 
33.6 
68.8 
59.0 
46.2 
45.5 

38.4 
49.0 
47.1 
56.5 
50.8 
24.7 
56.3 
47.7 
57.8 
73.4 
71.1 

52.1 
49.5 
51.7 
42.2 
48.1 



7o 
70.5 

72.6 
84.1 
63.4 
94.8 
97.2 
92.4 

92.7 
94.0 
82.5 
82.3 
90.2 

81.9 
74.2 
85.0 
80.5 
65.3 

59.1 
53.6 
66.4 
31.2 
41.0 
53.8 
54.5 

61.6 
51.0 
52.9 
43.5 
49.2 
75.3 
43.7 
52.3 
42.2 
25.6 
28.9 

47.9 
50.5 
48.3 
57.8 
51.9 



An Approach to the Study of Changes in Distribution / 305 



State 



Mountain 

Montana 

Idaho 

Wyoming 

Colorado 

Utah 

Far West 
Washington 
Oregon 
California 

Nevada 

* Above or below "All Employment" in Table 130. 
Source: Appendix Table Fl. 



Slow-growth 


Rapid-growth 


% 


% 


48.6 


51.4 


52.1 


47.9 


43.8 


56.2 


34.6 


65.4 


33.4 


66.6 


27.7 


72.3 


31.8 


68.2 


18.2 


81.8 


32.8 


67.2 



with 70% for the nation as a whole. Michigan, Massachusetts, New Jersey, 
and Connecticut derived most of their net upward proportionality shifts from 
a specialization in manufacturing, which accounted for 40% or more of their 
employment, as compared with a national average of 23.5%. 6 

This, of course, constitutes only a superficial explanation of the propor- 
tionality effect. All we have done here is to identify the employment sectors in 
each state that are statistically associated with the proportionality effect. In 
order to arrive at more fundamental explanations we need to seek the answers 
to two questions. 

First, why have some sectors of the economy grown more rapidly than others? 
In finding answers to this question we must concern ourselves with national 
demand and supply trends affecting the major employment sectors. And these 
must be analyzed in terms of important changes in basic economic determinants 
— population, resources, technology, and institutions. Second, where are the 
rapid-growth and slow-growth sectors located and why? Here we are brought 
to a consideration of the nature of the region and the principles governing the 
regional distribution of economic activity. It is evident that to answer these 
questions, more detailed sectors of the economy must be examined. 

Consider next the differential effect and the kinds of factors that are relevant 
to an explanation of it. A region that shows a differential net upward or down- 
ward shift does so because its expansion rate for specific component activities 
is greater than, or less than, the average for the same activities in other regions. 
This presents a significant contrast with the proportionality effect. The pro- 
portionality effect raised this key question: Why do some employment sectors 
of the national economy expand more than others? The differential effect 
raises another key question: Why does the same employment sector expand 

G The reader interested in viewing in greater detail the employment sectors in each state 
most responsible for the observed shifts can refer to Appendix Tables Fl and F2. 



306 / Regional Distribution of Economic Activities, 1939-1954 

more rapidly in some regions than in others? An answer to this question re- 
quires some knowledge of the reasons why one region may, for a particular 
activity, have access to inputs and markets superior to that of another region. 
It requires some knowledge of the significance of these spatial input-output 
relationships for multiplier effects of the sort discussed earlier. 

There are, necessarily, important differences in focus when the research design 
is applied to the differential effect as contrasted with the proportionality effect. 
In connection with the proportionality effect a region may experience a net 
downward shift because its relative access to markets has declined. The reason, 
however, is associated with the fact that the total market for its specialized 
sectors has declined relative to other sectors so that its share of the total market 
for all products and services has declined. At the first level the analysis con- 
cerns itself with the national trends in supply and demand. In connection with 
the differential effect, a region may likewise experience a net downward shift 
because its relative access to markets has declined. The reason in this case 
derives from the fact that its market access for a specific activity has declined 
relative to that exhibited by other regions engaged in the same activity. 

This makes it evident that much useful analysis of the proportionality effect 
can be accomplished through an investigation into national demand and supply 
factors and trends without consideration of the nature of regions other than 
to identify the locations of rapid-growth and slow-growth industries. However, 
little or no useful analysis of the differential effect is possible without taking 
into account modifications in regional input-output associations that represent 
changes in access for specific activities in specific regions. Such an analysis 
must aim at an understanding of the nature of regions on a basis that disaggre- 
gates space as well as employment activities. This is what we attempt to do 
for the years 1939-1954 in the chapters that follow. Our hope is that, through 
an analysis of specific economic activities in specific regions, we can clarify 
the working of the forces that are critical in influencing shifts in the relative 
volume of regional economic activities. 7 



7 However, the reader should be cautioned not to expect too much. Properly applied, this 
research design should be undertaken for a series to time periods. Our resources allow us 
only a glance at a single period — 1939-54. As a consequence, we can do very little in 
analyzing and generalizing about growth sequences and multiplier effects. All of the input 
and market ties that are important in defining the dominant access characteristics of 
specific activities in each region should also be analyzed. We will make a few gestures in 
this direction, but, even if adequate materials for doing this were readily available, the 
job is too monumental for more than partial realization here. We do, however, provide case 
materials that will be of value to persons concerned with problems of regional economic 
growth and we also present and illustrate a research design which we feel has considerable 
potential for regional research. 




Regional distribution of mining activity 1 



Mining is the smallest of the ten broad industry sectors, accounting in 1939 
for only 2% of total national employment. However, its significance for 
regional analysis is greater than its share in total employment would imply, 
for it provides a large part of the basic material inputs and thus is an origin 
for many complex sequences of production. For another reason, also, it has 
great significance for regional analysis. Probably no other economic activity 
is more unequally distributed geographically. Thus, in spite of its relatively 
small place in the national employment figures, mining can be most important 
in explaining the economic behavior of specific regions, and even more of 
specific localities. 

Mining employment in 1954 amounted to 754,238 persons as compared to 
827,410 in 1939 — a decline of 8.8%. The net shifts over these years amounted 
to ±165,390 workers, or 20% of the 1939 mining employment figure. In 
other words, while there was relatively little change in total employment over 
the period, there was a significant change in position among the states. In 
fact, the net shifts in mining employment were a larger proportion of the 
absolute changes than in any other of the major employment sectors. 

Table 132 gives a rough indication of how widely the major categories of 
mining departed from the national norms. The experience of the individual 
states in metal-mining employment, for example, was very different from the 
change in metal-mining employment for the nation as a whole (a mere .2% 
between 1939 and 1954) . By contrast, while total employment in coal-mining 
declined by one-third over this period, the net shift was fairly small, amounting 
to only 12% of the absolute change. The largest net shift in terms of numbers 
involved was in the petroleum industry (Col. 9) . 

^he term "mining" is employed in the sense in which it appears in the Standard In- 
dustrial Classification Manual of the Bureau of the Budget (1957), p. 21: "Mining is here 
used in the broad sense to include the extraction of minerals occurring naturally: solids, 
such as coal and ore; liquids, such as crude petroleum; and gases, such as natural gas. [It] 
is also used in the broad sense to include quarrying, well operation, milling (crushing, 
screening, washing, flotation, etc.) , and other preparation needed to render the material 
marketable. Exploration and development of mineral properties are included. Services per- 
formed on a contract, fee, or other basis in the development of mineral properties are 
classified separately but within this division." 

Wherever the term "mining" is used in this chapter, it may be taken to have this mean- 
ing except where the context clearly indicates a more restricted sense of the word. 

/ 307 



308 / Regional Distribution of Economic Activities, 1939-1954 



cs 

I— I 

CO 

Cs 









os -th © _; 
w o 2 Q 



00 p-5 *S __; 






«3 i— J 
O 



^ 






^ 



te 'Tl to 
-Q ° On 



^ 



ft 

E 
w 



^ 






p 00 On 


1— 1 r- , 


O 00 LO 

CM rH 


CM* ON 

"* i— 1 



o 


CO 


<ro 


LO 


<* 


CM 
CSI 


LO 

CO 


cm 


LO 


<* 

o 



© 


CO 


o 


<* 


*# 


On 


r- 


CO 


OS 


LO 


CO 


°l 


"*.. 


00 














LO 


t> 


00 


CO 


CO 


VO 


i-H 


CSI 


vO 





+1 +! +1 +! +1 



CO CM O 

CO ' CO 



CSJ I— I V© 

r- os ^h 
i— ^ o i— i 
co" csf csp 
r- eo 
csj 



8 



VO LO 

pH CO 



ft. LO 



"<* CO 

csi od 

On i-i 



CO CO 
"=f 00 

"* CO 



CO 00 


p— 1 


t- CM 


CO LO 


CM 


oo r- 


CM 00,, 


<* 


CO LO 








Tf< t> 




CM CM 


LO OS 


LO 


OS <-H 


t- 


CM 


CM r-t 



<# LO 

oo i-5 



LO l—i r-H 



CO CO 

i— T lo" 

LO On 



b 

a 
>-. 

o 

"ft 

G ' S 

■3 S 



£ 

i—; cfl 
P bfl 



B 93 

s a 

— < «j 5 s 

< & U V 



3 * 

03 ^ 



Regional Distribution of Mining Activity / 309 



Over-all Proportionality Effect 

Fifteen states showed a net downward shift because of their specialization 
in mining. Table 133 lists these states and indicates the proportion of total 
employment in each absorbed by mining. 2 

The key fact here is that basic mineral inputs have declined in relative im- 
portance when measured against the aggregate of all primary and intermediate 
inputs. The effect on employment is compounded by the substantial increases 
in productivity per man-hour, due mainly to mechanization, which have taken 
place in almost all mining industries in recent decades. Also, with the increasing 
specialization of functions, some functions that formerly belonged to the mining 



Table 133. Mining employment as percentage of total employment, 1939, for 
states with net downward proportionality shifts in total employment, 1939- 
1954*, because of specialization in mining. 

State 1939 

% 

United States average 2.0 



West Virginia 20.9 

Nevada 13.7 

Wyoming 7.8 4 

Kentucky 7.5 

Arizona 7.4 

Utah 7.2 

Montana 6.2 

New Mexico 5.7 

Oklahoma 5.3 

Colorado 4.6 

Alabama 3.2 

Idaho 3.0 

Kansas 2.7 

Texas 2.6 

Virginia 2.5 

*This is the total proportionality effect based upon 1-digit components shown in Figure 47. 

Source: Appendix Table Fl. 



industry have been split away. Specialized engineering and financial services, 
formerly provided by the mining industry, have now been absorbed by the 
business services sector of the economy. Some beneficiation processes formerly 

2 In addition to the states shown in Table 133, Pennsylvania (which had 7% of its em- 
ployed labor force in mining in 1939) experienced a downward proportionality effect from 
mining, but over-all had a net upward proportionality shift for the period 1939-1954 because 
it specialized mostly in rapid-growth sectors. 



310 / Regional Distribution of Economic Activities, 1939—1954 

included in mining have been segmented to the point that they are now classi- 
fied as manufacturing. 

Until the decade preceding 1920 the rate of growth in the physical volume of 
mineral production exceeded the rate for the gross national product. Since 
that time, however, it has been well below the gross national product rate. 
And the growth that has occurred, as well as the major share of the increase 
in productivity, has been largely contributed by oil and gas. Except in oil and 
gas, changes in mining output and productivity have been moderate, resemb- 
ling the changes in agriculture more closely than those in manufacturing. 
Borenstein 3 suggests that the reversal in the ratio of mineral production to 
national product is due, in part, to the predominance of mineral-saving over 
mineral-demanding developments in technology since World War I. Other 
factors have also contributed to the decline in the demand for minerals. For 
example, the more extended processing and fabrication which minerals undergo 
before reaching the final consumer generally adds to aggregate output without 
a commensurate increase in mineral input. There has also been an increasing 
use of scrap metals and of substitutes for minerals, such as plastics, rubber 
products, and laminated beams. 

In short, because of significant long-term trends in supply and demand fac- 
tors affecting minerals as a whole, the relative importance of the minerals 
sector of the economy has declined. As a result, the mining specialization states 
display net downward proportionality shifts in total employment (based upon 
1-digit components). 

There may be two factors at work, however, which can mitigate this in- 
fluence: (1) A state may combine good access to mineral resources with good 
access to markets and basic intermediate inputs. In Pennsylvania, for example, 
these offsetting factors generated a net upward proportionality shift for the 
period 1939-1954. (2) A region's access to mineral resources may be so good 
and/or may have improved so much relative to that of other regions that this 
may serve as the basis for net upward differential shifts in employment. In 
such a case, mineral resources may be a net stimulus to regional growth. 



Regional Influence of Mining Reflected in the Total Differential Effect 

In Figure 46 it was shown that some three-fourths of the differential net 
shift in total employment occurred west of the Mississippi. Here we are in- 
terested in examining the contribution of mining to the total picture. Figure 
48 shows the net shift in mining employment between 1939 and 1954. The 
states west of the Mississippi, it will be seen, contributed some 90% of the 
net upward shift; all but five of these Western states participated in this rela- 

3 For detailed documentation see Israel Borenstein, Capital and Output Trends in Mining 
Industries 1870-1948, Occasional Paper No. 45, (New York: National Bureau of Economic 
Research, 1954), pp. 10-28. Also, Barger & Schurr, The Mining Industries, 1899-1939: A 
Study of Output, Employment and Productivity (New York: National Bureau of Economic 
Research, 1944), pp. 3-58. 



Regional Distribution of Mining Activity / 311 

tive gain. Apparently the general regional picture for mining is roughly con- 
sistent with the dominant westward shift of total employment. 

Four other general observations of special significance emerge from a more 
detailed examination of the shifts. (1) The broad correlation alluded to above 
is subject to many individual exceptions. In about one-third of the states, the 
mining shift ran counter to the prevailing net differential shift at the 1-digit 
level for example, in Indiana, Ohio, Kentucky, Mississippi, and New York. (2) 
It cannot be claimed that, over-all, the shift in mining employment has been 
an important part of the total differential net shift in employment. There were 
only eleven states in which it amounted to more than 5% of the aggregate 




[ | Absolute upward shift 

i /A Relative upward shift 
I | Absolute downward shift 



Figure 48. Net Shift in Mining Employment, 1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 226.0. Total net shift as % of 1939 mining employment: 
20.0. Source: Appendix Table H. 



shift (both upward and downward) recorded by all ten industry sectors for 
the state. These eleven states are identified in Table 134. The principal in- 
fluence of mining upon the total employment shift is restricted to these states. 
For them, however, mining tells an important part of the story. As a matter 
of fact, the total shift in mining employment was about 20% of the total 1939 
mining employment. Thus, in relative terms, the shifts across state lines for 
mining were almost three times as important as for total employment. (3) Of 
these eleven states, those west of the Mississippi show absolute net upward 
shifts and those in the East — the Appalachian states — show absolute net down- 
ward shifts. This results, of course, from the substantial effect of the substitu- 



312 / Regional Distribution of Economic Activities, 1939-1954 

tion of oil for coal. Two-thirds of the net upward shift occurred in Texas and 
Louisiana and almost three-fourths of the net downward shift in Pennsylvania, 
West Virginia, and Kentucky. (4) A comparison of the mining shifts with the 



Table 134. States with relatively largest net differential shifts in mining em- 
ployment, 1939-1954. 



State Mining shift in state as % of total 

shift (without regard to sign) of 
1-digit employment categories for 
the state* 

Louisiana 36.3% 

West Virginia 24.4 

Wyoming 21.9 

Pennsylvania 15.4 

Kentucky 14.5 

Texas 14.5 

Oklahoma 10.9 

New Mexico 10.4 

Alabama 6.3 

Minnesota 6.3 

Utah 6.1 

*To make any fair comparison, it is necessary to sum these without regard to sign because 
the total net differential shift (i.e., the sum with regard to sign) balances out so that mining 
as a percentage of the net might be an extremely large figure in one or two states. 

Source: Table 157. 



total net differential shift (Figure 46) reveals that the net upward shifts in 
mining employment support those in total employment west of the Mississippi. 
In the East, however, the Upper South and the western end of the Manufactur- 
ing Belt show net upward total differential shifts in spite of absolute down- 
ward shifts in mining employment. This is especially surprising in the case 
of Kentucky, as mining was an important part of the changing picture for that 
state. 



1 -Digit Proportionality and Differential Effect in Mining Employment 
Based upon 2-Digit Components 

Table 135 shows the changes taking place in the four components of mining 
employment — coal, petroleum and natural gas, metallic minerals, and non- 
metallic minerals (other than organic fuels). At this level of aggregation the 
same proportionality and differential effects can be observed as in the shifts 
in total employment. The sum of the net shifts in the component sectors of 
mining rarely exhausts for a state the total net shift in mining employment 



Regional Distribution of Mining Activity / 313 

displayed in Figure 48. The balance is a proportionality shift resulting from 
the fact that some states specialize more than others in the rapid-growth or 
slow-growth sectors of mining. The proportionality and differential effects of 
the mining shift are presented in Figures 49 and 50. 




Upword shift 
| J Downward shift 



Figure 49. Differential Net Shift in All Mining Employment, 1939-54. 




Figure 50. Proportionality Net Shift in All Mining Employment, 1939-54. 



314 / Regional Distribution of Economic Activities, 1939-1954 

The proportionality effect is a more important part of the shift in mining 
(roughly 60%) than in total employment (roughly 50%). As Figure 50 shows, 
the proportionality net downward shift is confined to sixteen states. As a matter 
of fact, more than 90% of this shift occurred in the states along the Appalachian 
chain. Coal is, of course, the principal factor; employment in coal-mining de- 
clined sharply between 1939 and 1954. The three leading coal states — Pennsyl- 
vania, West Virginia, and Kentucky — account for 78.4% of the net downward 
proportionality shift. The dramatic growth of petroleum and natural gas is 
reflected in the fact that four leading petroleum states — California, Oklahoma, 
Texas, and Louisiana — account for 78.4% of the net upward proportionality 
shift. 



Table 135. Percentage change in mining employment in the United States by 
2-digit industry group, 1939-1954. 

Petroleum and natural gas +92.44% 

Non-metallic mining +18.27 

Metallic mining + 2.18 

All Mining Employment — 8.84 

Coal mining —48.01 

Source: Appendix Table G. 



The number of states showing net upward shifts because of a truly differential 
effect is much smaller. Twenty-seven states experienced net upward shifts in 
total mining employment. Seventeen of these displayed for their mining 
specialties a rate of growth that exceeded the national average for those cate- 
gories. These states fell into four groups — the Upper South, the Southwestern 
states (except Oklahoma), the easternmost Mountain states, and the northern 
Plains states. Two states, Texas and Louisiana, dominated the net upward 
differential shift during the period 1939-1954. Three states — California, Okla- 
homa, and Pennsylvania — dominated the net downward shift. An explanation 
of these shifts must be sought in the behavior of the individual mining seg- 
ments. 



Analysis of the Differential Effect in Mining 

In analyzing the differential effect, we are automatically concerned with the 
factors that lead the same industry sector of the economy to develop at different 
rates in different regions. Here the question of access to inputs and markets is 
of major importance. 

Mineral "inputs" are immobile inputs, and it goes without saying that min- 
ing activity can take place only where the minerals exist. In fact, no sector 
of economic activity shows a greater degree of concentration and localization. 



Regional Distribution of Mining Activity / 315 

Table 136 offers indication of the degree of concentration in mining ac- 
tivity. 4 Mining employment has a significantly smaller geographical associa- 
tion with population than any other major employment sector. The areas of 
mining specialization fall into three groups — the Mountain states, the South- 
west states, and the middle Appalachian states; in addition, three individual 
states have significant mining activities — Minnesota, Alabama, and South 
Dakota. In 1954 these states employed 75% of the total mining labor force 
and contained less than 30% of the total population of the country. 

It is obvious that resource access must be an important, if not dominant, 



Table 136. Rank correlations of each state's share of employment in selected 
industries with each state's share of population, 1954. 

Total employment .991 

Retail trade employment .979 Transportation and public 

utility employment .957 

Service & miscellaneous Finance, insurance & real 

employment .972 estate employment .957 

Wholesale trade employment .970 Manufacturing employment .936 

Government employment .970 Agricultural employment .649 

Construction employment .960 Mining employment .406 

Source: See Table 1, note and Appendix Table F-2. 



factor in explaining the behavior of the mining sector of the economy. But 
there are significant variations on this central fact. The extent to which specific 
resource deposits can establish dominance in the geographic pattern of economic 
activities is considerably modified by the existence of numerous substitution 
possibilities. 5 

The range of possibilities is great. At one extreme — admittedly hypothetical 
— is the resource for which there are no substitutes at all, either in terms of 
alternate sources or in terms of production technology. In such a case, extrac- 

4 Although this table is presented here in the mining section and alluded to only briefly 
at this point, it is a general reference table that we will refer to on a number of occasions as 
we develop this analysis. 

The term "substitution" is used here in its most general sense — the process of replacing 
one form of an input with another because of relative price relationships. At a more par- 
ticular level several types of substitution effects can be identified: important here are 
"source" substitutions in which a multiple-source commodity from a source developing lower 
"mill" prices will replace one from a source with higher prices. There are "technological" 
substitutions in which an input of different technical characteristics may replace an existing 
input when relative prices of the two differing inputs makes such a change advantageous. 
A recent example is the production and marketing of an aluminum "tin" can in competition 
with the traditional terne plate (coated steel) "tin" can. The phenomenon of "factor" sub- 
stitution is well known; for example, the extent to which an industry will substitute labor- 
saving machinery for labor inputs depends on interest rate-wage rate relations. 



316 / Regional Distribution of Economic Activities, 1939—1954 

tive activity would be unequivocally wedded to one location — the site of the 
single deposit. At the other extreme there are the resource deposits that are 
ubiquitous — for example, the mining of construction sand can utilize supply 
sources found in every region and almost every locality. In such cases 
market access rather than resource access is likely to determine the location 
of extractive activity; the ubiquity of resource deposits offers virtually an in- 
finite range of choices. 

Most situations are obviously intermediate. For example, there is a fairly 
wide geographic distribution of coal and iron ore deposits in the United States 
an$ thus a range of possible sites for extractive activity. In such cases, loca- 
tional decisions will be based upon the relative quality of these deposits and 
their accessibility as reflected in interregional price differences. These price 
differences, arising out of interregional differences in costs of extraction and of 
transport, will be the allocating factor determining (1) which of a series of 
resource deposits will be exploited, (2) the extent or intensity of its exploitation, 
and (3) the sequence in which competing sources will be brought into use. 6 

Sites of substitute resources offer a second set of substitution possibilities 
(for example, hematite vs. taconite vs. jasper; or coal vs. oil). In such cases 
the choice of location must admit the interregional price ratios of unlike, or 
substitute, resources as well as like resources. 

This series of choices between sites, both of like and unlike resources, deter- 
mines the geographic location of mining activities. 

In short, to undertake analysis of shifts in mining employment requires as 
much attention to all the factors influencing market and input access as the 
analysis of any other sector of economic activity. In the following section we 
want to examine the performance of the component sectors of mining employ- 
ment with this framework in mind. 

Regional Influence of Mining 

The mineral fuels (coal, crude oil, and natural gas) dominate the minerals 
resources picture. They account for roughly three-fourths of mining employ- 
ment and about three-fourths of the value of all mineral products. 

COAL MINING 

Influence of Coal-Mining in the 1 -Digit Proportionality Effect Exhibited by 
all Mining Employment: In 1939, coal-mining comprised almost 60% of all 
mining employment; by 1954 its share had dropped to only one-third, and oil 
and natural gas had taken over a large part of the potential as well as the 

"Further elements of the location choice arise out of the competition of uses for the same 
resource deposits. In such cases, interregional price differences have the effect of allocating 
favored sites to the uses that can pay the highest price. This competition of uses determines 
a fourth regional (and local) dimension — the distribution of resources to different uses and, 
therefore, the geographical distribution of these different uses. This is not an important 
element if the picture of extraction always forms an independent production stage. However, 
in cases where extraction is incidental to a resource-oriented processing stage, this can be 
important. It is also important in the distribution of site land and agricultural land. 



Regional Distribution of Mining Activity / 317 

conventional markets of coal. Over these years, the share of coal in meeting 
the nation's total energy requirements dropped from 51.4% to 29.3% (of 
total BTU's). At the same time the share of petroleum and natural gas in- 
creased from 44.9% to 66.8%. In many uses this economical energy source 
was a direct substitute for coal. For example, the dieselization of locomotives 
reduced the railway's use of coal from 120 million tons to 17 million tons be- 
tween 1945 and 1955. Oil and natural gas have also claimed a large share 
of the heating market and the electric utility market. While the absolute amount 
of coal used in generating electric power increased substantially during the 
period, the share of coal in this important market dropped from 81.7% in 
1945 to 65.6% in 1955 (and down to 55% in 1957). 




Relative upward shift 
| Absolute downward shift 



MM 

;:;:,.: I,:.., 



M'MlwMQl^SJ'LS. ! 



|W,V<3. Ky, %, UtqJi Ohk> &kL X<in. *'/ ■ _ _ A to. towo It*. Po. 

" ~~ .within these limits 



20 



40 






Figure 51. Net Shift in Coal Mining Employment, 1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 12.3. Total net shift as % of 1939 coal mining 
employment : 5.9. Source: Appendix Table I. 



Another important reason for the relative decline of coal in the energy market 
is that oil (and the new oil technology) generated new energy sources and 
energy markets (notably the internal combustion engine) in which coal was 
not a reasonable technological substitute. Some of the factors leading to a rela- 
tive decline in the importance of the mining sector of the economy as a whole 
might also have a differential effect upon coal-mining. For example, the tend- 
ency in recent years for mineral-saving developments to predominate over min- 
eral-demanding developments in technology seems to have been especially im- 



318 / Regional Distribution of Economic Activities, 1939-1954 

portant in coal technology. The average amount of coal burned to produce 
one kilowatt-hour of electrical energy in the United States was 6.4 pounds in 
1902, 3.0 pounds in 1920, and 1.2 pounds in 1950. 7 

Influence of Coal-Mining in the 1 -Digit Differential Effect Exhibited by all 
Mining Employment: The important role played by coal in the proportionality 
effect in the shift in mining employment does not mean that differential shifts 
between regions in coal-mining have not been an important element in the 
shift picture. In Figure 49, seventeen states were identified as having had net 
differential upward shifts in mining employment. The extent to which coal 
contributed to these shifts is shown in Figure 51. The net upward differential 
shift shown by the mid-Appalachian states in Figure 49 is plainly due in large 
part to a relatively strong net upward shift in coal-mining employment for 
those states. As a matter of fact, Kentucky, West Virginia, and Virginia ac- 
count for almost 90% of the net relative upward shift. 8 Utah also gains at 
least a part of its differential strength from coal-mining. On the other hand, 
Pennsylvania, Illinois, and Alabama obviously derive an important part of 
their differential net downward shift in mining employment from a net down- 
ward shift in their share of coal-mining employment. 

A relatively small number of states are involved in these shifts because of 
the high degree of regional concentration of coal production. While some coal 
reserves are found in thirty-three states and some coal production takes place 
in twenty-six, the major portion of mining activity is heavily concentrated. The 
Appalachian field, extending from northwestern Pennsylvania to Alabama, 
dominates American coal production. Three states in this area — Pennsylvania, 
West Virginia, and Kentucky — produce 75% of the output. Elsewhere, only 
southern Illinois and the mid-Mountain states have any significant degree of 
concentration. 

The shifts noted above would seem to have come about through the workings 
of a number of forces. In a study based on an examination of 94 State Econ- 
omic Areas 9 producing coal, Beverly Duncan has reviewed the general access 
characteristics of the coal industry. 10 In this study she established that the 
inferior deposits, taken as a group, show a marked tendency to be exploited 
more intensively in areas of high market accessibility. She also took produc- 
tion as a ratio of available reserves and correlated it with population potential 
(i.e., closeness to the major centers of population) as an index of market access 
for the 94 coal-producing State Economic Areas. The coefficient was .65 and 
the correlation was significant at the .01 level. 

She concludes that population potential explains slightly more than two- 
fifths of the variation in the "index of resource exploitation." This finding, of 

7 Petroleum technology made some efficiency gains to be sure but these were concentrated 
more in the field of motor fuels where coal is non-competitive. 

8 In this case, "upward" shift means that the absolute losses in this area were less than 
elsewhere. 

9 These are groups of contiguous counties with similar economic and social characteristics. 

"Reported in Beverly Duncan, Population Distribution and Economic Activity: The Non- 
metropolitan United States in 1950, Ph.D. Dissertation, University of Chicago, 1957 (micro- 
film) , Chapter 3. See also O. D. Duncan, et al, Metropolis and Region, op. cit., Chapter 8. 



Regional Distribution of Mining Activity / 319 

course, is completely consistent with the hypothesis that the intensity with 
which coal resources are exploited increases as accessibility of the resource site 
to the national market increases. It should be pointed out here that the re- 
serves of higher-grade coal tend to be concentrated in or toward the relatively 
high potential areas. Insofar as the higher ranks are preferred for reasons 
other than or in addition to their heat content (which has been allowed foi 
here through the use of BTU equivalents), the direct relationship between in- 
tensity of production and population potential is strengthened by the areal 
distribution of mineral resources. In other words, the relatively high potential 
areas occupy a favorable position both from the standpoint of the quality of 
their deposits and their proximity to market. 

Figure 52 gives some idea of the extent to which the heavy concentrations 
of industry and population in the northeastern part of the country affect 
intensity of coal production. While some of the most extensive coal reserves 
are located in the Mountain states, (particularly Montana, Wyoming, Utah, 
and Colorado), these states, with few exceptions, have subareas falling in the 
lowest quintal of the production-to-reserve ratio. The intensity of exploitation 
of coal reserves is highest in the Appalachian Coal Plateau region, with some 
State Economic Areas in southern Illinois and the western parts of Kentucky 
also showing a high rate of exploitation. No SEA west of the Mississippi River 
falls in the highest quintals of the production-to-reserve ratio. 11 

Small-area data are available on the value received or charged for coal f.o.b. 
the mine, including the selling cost. 12 An analysis of these data reveals no 
significant relationship between f.o.b. value per BTU produced and population 
potential, a summary measure of access to the national market; there is only a 

"Estimates of coal resources by rank (lignitic, sub-bituminous bituminous, and semi- 
anthracitic) at the state level have been prepared and published by the U. S. Geological 
Survey; these resources were distributed among SEA's. The estimates are not altogether 
comparable on an inter-state basis; and, of course, the allocation of state reserves to SEA's 
was quite arbitrary in some cases. Nonetheless, a quantitative indicator (tonnage) of 
remaining coal resources by rank is available for each SEA. To obtain a single quantitative 
indicator of all remaining coal resources for each SEA, the tonnage for each rank of coal 
was converted to British thermal unit (BTU) equivalents, and the respective equivalents 
then aggregated into a single statistic. The conversion factors employed assume 26.0 
million BTU's per ton of bituminous coal, 25.4 per ton of anthracite, 19.0 per ton of sub- 
bituminous coal, and 13.4 per ton of lignite. See Coal Resources of the United States, by 
Paul Averitt, Louise R. Berryhill, & Dorothy A. Taylor, Geological Survey Circular 293, 
(Washington, 1953), pp. 6-12. 

Small-area statistics on coal production are available on an annual basis from the Minerals 
Yearbook, prepared by the U. S. Bureau of Mines. Production is reported separately for 
lignite, bituminous coal, and anthracite. For the analysis, the tonnage of bituminous coal 
produced was divided into sub-bituminous and bituminous coal in proportion to reserves. 
This is a highly arbitrary procedure, but at a maximum it affects only the 16 SEA's in which 
reserves of both sub-bituminous and bituminous coal occur. 

12 The producer's estimate of the value of coal not sold but used by him also is included. 
The value figures, then, do not include charges incurred in transporting the coal from the 
producer to the purchaser. [U. S. Bureau of Mines, Minerals Yearbook 1950 (Washington: 
Government Printing Office, 1953, p. 322.] Perhaps the most conventional "value" 
indicator is value per ton produced; but areal variation in value per ton is likely to reflect 



320 / Regional Distribution of Economic Activities, 1939-1954 




Regional Distribution of Mining Activity / 321 

tenuous inverse relationship between f.o.b. value per BTU and population po- 
tential (r= — .10). 

Cursory inspection of statistics on bituminous coal and lignite distribution 
suggests that the absence of any positive association between f.o.b. value per 
BTU and population potential may "explain" in part the relatively low "index 
of resource exploitation" values observed in the "low-potential" areas. This 
proposition is, however, sheerly speculative, for no rigorous analysis can be 
undertaken with the data available. However, the destination of shipments of 
coal produced in the "low-potential" areas, insofar as it is known, was generally 
within the area near the site of production. For example, 94% of the coal 
produced in Colorado was shipped within the Mountain area (Census Division) 
in which Colorado lies; 60% of the coal produced in Utah was shipped within 
the Mountain Division in which it lies and 39% to the adjacent Pacific Division. 
All shipments of the lignite produced in North and South Dakota were within 
the West North Central Division of which these states form a part. 13 The 
statistics also show that these divisions were the destination of relatively few 
shipments from other areas; in other words, it appears that divisional pro- 
duction at the present level is sufficient to meet most divisional needs. With 
respect to the "national" market, the areas of low population potential clearly 
have a disadvantageous competitive position, for the costs they incur in trans- 
porting coal to the "national" market are greater than those incurred for coal 
produced in "high potential" areas. In the absence of a differential f.o.b. value 
favoring the "low potential" areas, their production seemingly must be geared 
to "local" or "regional" needs. Insofar as f.o.b. value per BTU and the "index 
of resource exploitation" are independent of one another on an area-by-area 
basis (r = .05), it is suggested, speculatively to be sure, that intensity of pro- 
duction is associated with population potential, or ease of access to the national 
market, through differential transport costs. 14 

Here we have an industry in which immobile resource inputs make resource 

in large part simply areal variation in the rank of coal produced. Furthermore, any attempt 
to evaluate value per ton in relation to population potential is especially difficult, for anthra- 
citic and bituminous coal reserves tend to be concentrated in or toward the high-potential 
areas whereas the reserves of sub-bituminous and lignitic coal tend to be concentrated in 
the low-potential areas. Consequently, an alternative "value" indicator which has a built-in, 
rough control on the "quality" of coal produced — value per BTU produced — was employed. 

13 U. S. Department of Commerce, Bureau of the Census, "Bituminous Coal and Lignite 
Distribution," 1954 Census of Mineral Industries, Vol. I (Washington: Government Printing 
Office, 1958), Section 12B. 

"The analyses reported above are of an exploratory character. The dearth of small-area 
statistics on mining activity and mineral resources hampers any investigation of the patterns 
of areal variation in mining. To some extent, the forthcoming release of the detailed results 
of the 1954 Census of Mineral Industries should permit a more thorough investigation of the 
relationships discerned here. 

Figure 52 — opposite. Tons of Coal Produced per Million Tons of Reserves, 
Measured in British Thermal Units, 1949, by State Economic Areas. 

Source: An unpublished series of data for state economic areas prepared 
by Donald J. Bogue and Calvin L. Beale. 



322 / Regional Distribution of Economic Activities, 1939-1954 

access a major factor in the regional distribution of activity. On the other 
hand, there are sufficient alternative sources of the same input to make market 
access an important consideration in the distribution of the industry. Further, 
the existence of a major substitute product can modify the dimensions of market 
access considerably. 

The interregional shifts shown in Figure 51 are directly tied to changes in 
these access factors during the period under examination. 

Consider, first, factors that may have changed the conditions of input access. 
Resource depletion has undoubtedly exerted some influence. The shift out of 
Pennsylvania into the mid-Appalachian fields has arisen in part from the fact 
that many of the most productive seams have been exploited to the point where 
mid-Appalachian seams formerly of lower quality have become competitive. 15 

Mechanization of coal mining has also had an important influence. The 
production economies that have resulted from mechanization in the areas that 
are adaptable to it 16 have depressed many other coal-mining regions below 
the margin of feasible economic exploitation. This seems to be an important 
factor in the major increase in the share of production in the mid-Appalachian 
fields and the decline in importance of most of the areas outside. Mechaniza- 
tion has altered the comparative advantage of coal deposits in such a way that 
many deposits, in effect, have been redefined out of existence as an economic 
resource, given present demand and technology; and for many other deposits, 
profitable levels of exploitation have been sharply modified. 

A third factor may have been important. Labor represents 76% of the pur- 
chased inputs for the coal industry. It is possible that institutional factors, such 
as strong unionization of the coal miners, may have contributed to a change in 
relative regional access to inputs in the sense that rising labor costs may, in 
some fields, have submerged production below the margin of economic ex- 
ploitation. This could be a contributing factor leading to a maximizing of 
production in those fields where output per man-hour tends to be the greatest. 

The combined influence of these factors affecting resource access seems to 
have increased the importance of the mid-Appalachian states relative to other 
coal-producing regions. 

In a similar way interregional changes in relative access to markets has 

^This might imply either lower physical quality and equal market accessibility or it might 
mean equal physical quality with unequal access to the prime markets or both. 

Empirical studies indicates that the decline of coal mining in Pennsylvania might have 
been even more excessive had not open-pit and augur mining methods been substituted for 
underground mining. In 1940, only about 3 per cent of the state's soft coal production was 
mined by surface methods; by 1953, over 21 per cent was so mined. Total reserves of 
bituminous coal in Pennsylvania however, are still extensive. See especially: George F. 
Deasy and Phyllis R. Griess, "Geographical Significance of Recent Changes in Mining the 
Bituminous Coal Fields of Pennsylvania." Economic Geography. Vol. 33 (October 1957), 
pp. 283-98. 

16 Effective use of both underground and surface mining equipment requires seams without 
too many faults. Moreover, underground mining requires seams that ideally are 4-6 feet 
thick (although lesser and greater thicknesses are exploited) ; and surface mining requires 
seams at comparatively shallow depths. The western fields, in particular, in many instances 
tend to be marginal from the point of view of both thickness of seams and access to markets. 



Regional Distribution of Mining Activity / 323 

influenced the shifts we observe. The extent to which the substitution of oil 
for coal in the power market has been regionally concentrated is shown in 
Figure 53. In general, oil dominates the power market west of the Mississippi 
and coal east of the Mississippi. Thus coal has succeeded in holding this market 
in the areas with greatest access to prime coal sources and has lost out in the 
areas with greatest access to oil. The fact that the substitution of oil for coal is 




ESa 75 % or more supplied by coal 

1 '.:-:■'■ I 75% or more supplied by 

oil or natural gas 
H Mixed 

Upper number is percentage of coal used 
NA — not available 



Figure 53. Relative Use of Coal and of Oil or Natural Gas by Steam Electric 
Plants, 1955, by State. 

Source: National Coal Association, Department of Coal Economics. 



not equally feasible in every region is another element serving to explain the 
relative gains of coal production in the mid-Appalachian area. This region 
was in the strongest competitive position for retaining a larger share of the 
declining coal market. 17 

The eastern coal areas were less vulnerable to market loss through substitu- 
tion for another reason also. This is the area where the nation's manufacturing 
is concentrated, and it is in the manufacturing sectors that coal seems to have 
had the greatest success in retaining its market. 18 The portion of total output 
that goes to manufacturing is about three times greater for coal than for oil 
and gas. 

"The loss of markets in home-heating and transportation has exhibited a similar regional 
differentiation with similar effect. 

18 There is considerable logic to this. The 100% weight-losing characteristic of fuel can 
exert much more influence in manufacturing than in the other major coal markets. House- 
holds do not move to fuel to realize economies in heating; and the economics of power 
transmission require that electric utilities be close to markets, not fuel sources. Manu- 



324 / Regional Distribution of Economic Activities, 1939-1954 

Table 137. Relative use of coal and coke and of oil and natural gas in manu- 
facturing in Pennsylvania, West Virginia, and Texas, by industry, 1947 

(per cent) 





Coal and coke 


Oil and 


natural 


gas 


2-digit 














manufacturing categories 


W. Va. 


Penna. 


Texas 


W. Va. 


Penna. 


Texas 


Food & kindred products 


99.2 


83.5 


59.8 


.8 


16.5 


39.4 


Tobacco manufacturing 


— 


69.8 


— 


— 


30.2 


— 


Textile mill products 


99.7 


77.2 


29.3 


.3 


22.8 


70.6 


Apparel & related 


99.8 


79.3 


— 


.1 


20.7 


99.6 


Lumber & products 


82.7 


74.7 


28.2 


17.3 


25.1 


70.9 


Furniture & fixtures 


100.00 


89.8 


68.9 


— 


10.5 


31.4 


Paper & allied products 


100.00 


90.6 


— 


— 


9.4 


99.3 


Printing & publishing 


87.3 


82.2 


55.5 


12.6 


18.2 


43.4 


Chemical & allied products 


99.8 


82.7 


19.6 


.2 


17.3 


79.9 


Petroleum & coal products 


95.9 


86.8 


— 


3.2 


11.7 


99.7 


Rubber products 


— 


13.4 


— 


— 


17.3 


— 


Leather & leather products 


100.0 


95.5 


— 


— 


4.5 


99.5 


Stone, clay, & glass 


96.9 


91.0 


31.5 


3.1 


4.7 


68.3 


Primary metals products 


91.7 


89.6 


82.2 


8.1 


10.3 


17.8 


Fabricated metals products 


99.3 


75.9 


44.5 


.7 


24.0 


55.1 


Machinery (exc. elec.) 


93.7 


82.2 


98.0 


6.3 


23.8 


1.9 


Electric machinery 


99.9 


93.0 


97.4 


— 


7.0 


2.2 


Transportation equipment 


— 


51.4 


2.1 


— 


12.4 


97.7 


Instruments & related 


— 


74.2 


— 


— 


23.4 


99.1 


Miscellaneous manufacturing 


97.8 


79.6 


67.5 


2.2 


19.1 


32.4 



Source: Leland W. McCloud, Comparative Costs of Competitive Fuels, A Study of Fuel 
Consumption by Manufacturing Industries in the United States, 1947, West Virginia 
University, Business and Economic Studies, Volume 1, June 1951, Tables 4 and 5. 



Table 137 indicates the considerable degree to which coal is successful in 
retaining its markets in manufacturing even in areas where its competitive 
position is weakest. 19 In Texas twelve of the twenty 2-digit manufacturing 

facturing industries have greater freedom in seeking and utilizing the cheapest fuel sources. 
In many instances this turns out to be coal. Moreover, the technology of manufacturing in 
many industries may not allow for easy substitution of oil for coal. 

Coal became the established industrial fuel before oil and gas came on the scene. It is 
no accident that the major manufacturing belt is located close to the nation's choicest coal 
sources. And agglomeration economies and market concentration have kept the manufactur- 
ing belt's share of employment and output from diminishing to any significant extent. 
Since most of the nation's manufacturing is conducted in this belt, close to coal's strongest 
competitive position, it is not too surprising that coal has succeeded in holding an im- 
portant portion of this market, in spite of improved transportation, improved fuel tech- 
niques, and the alternative of oil. 

19 Since the 2-digit groups are still gross aggregates, we are aware that this state-to-state 
comparison might be misleading because of differences in the composition of the industries 
making up these aggregates. The general picture, however, can be considered as valid. 



Regional Distribution of Mining Activity / 325 

groups used coal and coke for 20% or more of their energy requirements (and 
nine for more than 50%). In contrast, oil completely failed to penetrate the 
manufacturing markets in West Virginia and showed only moderate penetration 
in Pennsylvania despite the fact that pipe lines and ocean transport give Penn- 
sylvania favorable access to oil and gas fuels. Primary metals and machinery 
rely almost entirely on coal even in the heart of oil's competitive stronghold. 20 
Miscellaneous manufacturing and food products also continue to be strong 
coal users. 

One further change in the market has had an important effect on the regional 
changes in coal production and employment. A sizable part of the output of 
the coal industry goes to export (12.5% in 1947). Between 1939 and 1954 
these exports tripled. The fact that the coal fields with most favorable access 
to ocean shipping are located in West Virginia, helps to explain the strong 
relative showing of that state. 

We have described here an industry seriously affected by loss of markets 
and have shown some of the regional consequences. At this point, a note of 
caution might be appropriate against any assumption that the regional and 
national trends exhibited will necessarily continue. Energy is one of the 
fastest growing segments of the economy and the price trends of gas and oil 
are now starting to favor coal after working against it for thirty years. (Indeed 
public utility markets in New England, the Southeast, and the big metropolitan 
markets of New York City and Philadelphia have already started to return to 
coal.) Add to this the prospects that may develop out of hydrogenation, 
carbonization and gas synthesis, coal pipelines, etc., and it is possible that the 
prospects for coal in the years to come may improve. 



PETROLEUM AND NATURAL GAS PRODUCTION 

The degree of regional concentration and specialization in petroleum and 
natural gas production is not greatly different in general from that observed 
for coal. While twenty-four states produce some oil and gas, five states — Texas, 
Oklahoma, Louisiana, Kansas and California — employ about 78% of the national 
working force and produce 85% of the national output. The degree of popu- 
lation association (see Table 138, infra) is approximately the same as for coal. 

The regional distribution, of course, differs markedly. The Appalachian coal 
area, where oil and gas were first discovered, now supplies less than 5% of 
the total output. Instead, most production comes from the Southwest states 
plus Kansas and California. 

Petroleum and natural gas production have had a marked influence on the 
1 -digit proportionality shifts exhibited by all mining employment. As pointed 

2n Coal's retention of a major part of the market in primary metals, machinery, etc., helps 
explain the experience of Utah and Oklahoma in Figure 51. As population and economic 
activity have expanded in the West and Southwest, there has been an expansion in those 
manufacturing activities that maintain their allegiance to coal. Utah's strength in coal can 
probably be explained to a considerable degree by the establishment and expansion of steel 
facilities at Provo, Utah, and in Southern California. 



326 / Regional Distribution of Economic Activities, 1939-1954 

out earlier, four leading petroleum states (California, Oklahoma, Texas, and 
Louisiana), accounted for 78.4% of the net upward proportionality shift 
(Table 50). This important proportionality effect arises from the fact that, 
alone among mining industries, the petroleum and natural gas industry can 
claim to be a "growth" industry. 

But shifts among the states in total mining employment has also risen from 
the fact that some "oil" states have made out better than other "oil" states; that 
is, there has been a differential effect as well as a proportionality, or industry- 
mix, effect. If we examine once again the seventeen states that showed a net 
upward differential shift for all mining employment (Figure 49) and compare 
that with the net shifts in crude petroleum and natural gas employment, we 
find that 9 of the 17 states receive an important boost from upward shifts of 



Upwof d shift t 

50 
40 

; 30 

20 
10 



10 
20 
30 

Bownwwd shift* 



j Absolute upward shift 
J Relative downward shift 
_] Absolute downward shift 



Tex, Uu Co}o.Wyo.Mi$5.N.M 



L „J AH other statesji 




Figure 54. Net Shift in Employment in Crude Petroleum and Natural Gas 
Production, 1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 45.5. Total net shift as % of 1939 crude petroleum and 
natural gas employment: 42.1. Source: Appendix Table I. 



oil and gas employment. 21 In all of these except Utah petroleum dominates the 
shift in total mining employment. (Utah also gained from coal, as we saw in 
the previous section.) Figure 54 shows the states at the ends of the shift 
spectrum. 

"'Mississippi, Louisiana, Texas, New Mexico, Colorado, Utah, Wyoming, North Dakota, 
and Nebraska. 



Regional Distribution of Mining Activity / 327 

In some of the states the net downward differential shift in total mining 
employment has been due to relative losses in petroleum employment. The 
three states that show almost two-thirds of the net downward shift in total 
mining employment — California, Oklahoma, and Pennsylvania — also show two- 
thirds of the net downward shift in oil and gas employment. 

Informed opinion holds that the changes in regional positions here are tied 
to resource depletion and discovery and that markets and non-resource inputs 
are relatively unimportant in these shifts. 22 In general, the earliest discoveries 
were in the Northeast; the Oklahoma and Texas fields were brought in at an 
intermediate period, and the most recent additions to resource capacity have 
been in Louisiana, Texas and the eastern Mountain states. 23 The Northeastern 
reserves have been worked for so long that depletion has forced absolute 
declines in employment and production. California and Oklahoma are not 
suffering from depletion in the sense that there have been any absolute declines 
in production. However, their relative position has declined because of their 
failure to discover new reserves at a rate commensurate with that of Louisiana, 
Texas, and the eastern Mountain states. 

Oil and gas production, therefore, is unique among mining activities in that 
significant additions to reserves and well capacity have been made during the 
period under examination. 24 This factor, rather than changes in markets and 
in production and marketing technology, would seem to have dominated the 
regional shifts observed. 25 



METALLIC MINERAL MINING 

Employment in metal-mining showed a 2' 
as compared with an 8.8% decrease for mining as a whole. Thus, while spe- 
cialization in metal-mining had the effect of generating net upward propor- 
tionality shifts in mining employment, the tendency to do so was not very strong. 

However, it does serve to explain the net upward proportionality shift for 
all mining in those states that have important metal-mining activities. Produc- 
tion and employment are largely concentrated in the Mountain states and the 
glacial shield in the Great Lakes region. Elsewhere, only Alabama and the 

22 This is a consensus gleaned from the general literature and discussions with the oil and 
gas specialists at the Bureau of Mines and at Resources for the Future. 

23 There has been considerable overlapping of time periods, of course. 

24 To a considerable degree, new "additions" to reserves in this period may reflect (a) 
differences in reserve "bookkeeping" between coal and oil, and (b) differences in the 
nature of oil and coal "discovery." 

^However, v/hile the role of changing markets, technology, and institutions do not seem 
to have been central to the regional changes in the industry, it may well be that the loca- 
tion of drilling activity and rates of exploitation have been somewhat influenced by these 
factors. It may also be that institutional factors such as an increasing trend to vertical 
integration have encouraged new drilling in marginal areas. At this point we are not in a 
position to evaluate the relative importance of these factors but, clearly, they deserve further 
study if a comprehensive picture of all the forces that have been affecting the location of 
mining activities for mineral fuels is to be achieved. 



328 / Regional Distribution of Economic Activities, 1939-1954 

Ozark plateau show specialization. There are twenty states with little or no 
metal-mining activity, while the Mountain states, together with Minnesota and 
Michigan, account for about three-fourths of the activity. 

Figure 55 gives some idea of the contribution of metal-mining to the net 
differential shifts in mining employment. Since total employment in this mining 
sector remained so nearly stable, the shifts are represented as absolute upward 
shifts or absolute downward shifts. For the most part, the states east of the 
95th meridian show absolute gains in their shares of employment and the states 
west of it absolute losses. The major gains were made in the glacial shield 
region — Minnesota, Michigan, New York — and in Arizona. The biggest losses 
were in California, Oklahoma, and Alabama. 

The access characteristics of the regions engaged in metal-mining suggest 
some of the reasons for these developments. Metal-mining, more than any other 
mining activity, exploits an immobile input for which there are relatively few 
alternative sources within the United States. Thus, in general, a fruitful do- 
mestic source will tend to be exploited irrespective of market access considera- 
tions, as Table 138 suggests. The association with population and population- 
centered activities seems to be slighter in metal-mining than in any other of the 
economic activities we survey (resource or otherwise). Thus we can concen- 
trate our analysis of the regional access changes upon metal resource discovery 




[Ijlllf Absolute upward shift 

_J Absolute downward shift 

» No activity 



Figure 55. Net Shift in Metal Mining Employment, 1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 859.8. Total net shift as % of 1939 metal mining 
employment: 18.8. Source: Appendix Table I. 



Regional Distribution of Mining Activity / 329 

Table 138. Rank correlations of each state's share of employment in selected 
mining activities with each state's share of population, 1954 



Total mining employment .406 

Metal mining employment .001 

Coal mining employment .197 

Crude petroleum and natural gas employment .226 

Other non-metallic mineral employment .828 

Source: U. S. Bureau of the Census, Census of Mineral Industries, 1954 Preliminary Re- 
port, Series MI-14-3 (Washington: Government Printing Office, 1956) ; U. S. Bureau of 
the Census, Current Population Reports: Population Estimates, Series P-25, No. 145 
(Washington: Government Printing Office, 1956), Table 2, Page 6. 

and depletion and technological changes affecting the cost conditions of supply. 26 
The most obvious starting point for assessing total metal-mining employment 
is iron ore production, which is responsible for 44% of the total if precious 
metals are excluded. The importance of iron ore in the metals complex, to- 
gether with the fact that the nation's industrial growth has made greater 
demands on it than on any other metal, provides the principal first-level ex- 
planation of the eastward shift in metal-mining employment and the dominance 
of the glacial shield areas. 27 Minnesota, which shows 41% of the net upward 
shift in mining employment, shows 39% of the net upward shift in iron ore 
production. New York, Pennsylvania, and New Jersey show another 12% of 
the net upward shift. While Michigan and Wisconsin did not increase their 
share of production, their absolute increases were such that they show 11% 
of the net upward shift of total metal mining employment. Alabama, which 
shows 8% of the net downward shift for total metal mining, shows 31% of 
the net downward shift in iron ore production. 

Known iron reserves in the United States have not changed in our period, 
and up to 1954 depletion had not caused any serious curtailment of produc- 
tion (although it is now beginning to do so). The principal change in access 
therefore, must be associated with changes in the market for iron ore over 
these years. The most significant change here was the change in the size of 

26 A complicating element derives from the fact that metal mining is not a relatively 
homogeneous activity like coal mining or petroleum and natural gas production. It embraces 
the mining activity of some twenty or thirty metals. A part of the shift in metal mining 
employment should be analyzed, therefore, in terms of the proportionality effect of its 
component activities. This is not a complicated job, but at the time these materials were 
being brought together, the relevant minerals statistics for the terminal year of our period 
were not available. Because of the large number of different metals involved, comprehensive 
analysis of the regional modifications of access to metallic mineral inputs and markets is 
also beyond our scope. The best we can do here is to "explain" the shifts in metal mining 
employment at a superficial level by decomposing this change into the major metal con- 
stituents that make it up. We will go beneath this to the access characteristics only to the 
extent that we may be able to give broad hints. 

27 The production of iron ore during our period increased by 189% as compared with a 76% 
increase for all metallic minerals. 



330 / Regional Distribution of Economic Activities, 1939-1954 

the total market. The production of pig iron, and therefore the production of 
iron ore, roughly tripled during the period. Because of the regional concentra- 
tion of iron ore deposits, the net effect of the expansion in total demand was 
to establish the dominance of the Great Lakes states in total metal-mining dur- 
ing the 1939-54 period. 

There are some other regional effects of interest that the total metal-mining 
picture disguises. (1) The decline in metal-mining employment and iron ore 
production in Alabama is directly related to a decline in that state's share of 
pig iron production. (2) Because metal-mining other than iron ore is so im- 
portant in the West, the data for total metal-mining employment (shown in 
Figure 55) give no hint that California, Utah, and Texas account for 42% of 
the net upward shift in iron ore mining between 1939 and 1954. This is direct- 
ly related to the fact that the "undistributed" category in the Minerals Year- 
book — California, Colorado, Utah, Texas, Massachusetts, Minnesota, West Vir- 
ginia and Tennessee — accounts for 80% of the net upward shift in blast furnace 
output. Of these states California, Texas, and Utah dominate the increase. We 
make no serious attempt here to go deeper and identify those changes in access 
that have, in turn, resulted in this shift in blast furnace output. It seems evident, 
however, that the increases in population and industrial activity in the West and 
Southwest constitute an important market shift that tells an important part of 
the story. 

For the rest, let us make a quick run-down of the principal factors associated 
with the major shifts in metal-mining employment. California supplied 44% 
of the net downward shift in metal-mining during our period. This was due 
largely to the decline in production of gold and silver. 28 Here is a place where 
resource depletion seems to be a major element in the change in access. This 
has been an important element in the whole picture of the eastward shift of 
metal-mining employment. Among the states west of the 95th meridian, only 
Utah showed an absolute increase in gold and silver production — largely a by- 
product of copper mining. The rest recorded steep declines, led by California. 

Since almost 90% of Arizona's mining activity is associated with copper and 
Arizona produces 43% of all copper ores in the United States, its unique per- 
formance in the West can be tied directly to copper. During the 1939-54 period 
Arizona claimed 50% of the total net upward shift in copper-ore mining. The 
major factors affecting the regional picture in copper-mining seem to be a 
mixture of technological change and resource discovery. In recent times pros- 
pecting has added significantly to the known copper reserves in Arizona, New 
Mexico, and Utah. During the post-depression period huge new machines for 
open-pit mining have been developed. This has significantly altered the com- 
parative cost advantage of open-pit mines as compared with shaft mines. Both 
of these factors have favored Arizona at the expense of the older shaft mines 
in Michigan and Montana. 

The large net downward shift exhibited by Oklahoma and Kansas is tied to 

28 Actually, California had increases in many other metal activities including tungsten, 
chromite, iron ore, lead and zinc. These were not sufficiently important in the total to 
affect the impact of gold and silver upon the total. 



Regional Distribution of Mining Activity / 331 

declines in the output of lead and zinc in the Ozark plateau area. The wide- 
spread net upward shift in metal-mining employment in the eastern states out- 
side the Great Lakes region is associated with a wide variety of lesser minerals 
that are growing in importance in a metallurgical world: bauxite and man- 
ganese in Arkansas; titanium, zirconium, and uranium from phosphate ores 
in Florida; lithium and a new tungsten mine in North Carolina; and titanium, 
lithium, and manganese in Virginia are some of the growing metal mining ac- 
tivities in the East. 



NON-METALLIC MINERAL MINING 

Non-metallic mineral mining has exerted a stronger influence than metal- 
mining upon the proportionality effect exhibited by mining employment as a 
whole. Both, of course, are definitely subordinate to the mineral fuels in this 
respect. Since the area of greatest specialization in non-metallic minerals has 
been the East, it is here that it has made its greatest contribution to the net 
proportionality shift in total mining employment. 

Figure 56 suggests the contribution made by non-metallic minerals to the 
differential net shift in total mining employment. 29 On the whole, states west 
of the 90th meridian experienced net upward shifts, and states east of this line 
experienced net downward shifts. 

Non-metallic mineral mining employment has a closer association with 
population and its associated activities than any other resource activity (Table 
138). 30 The minerals that comprise this group provide an excellent example 
of the extreme case of substitutability mentioned earlier in the chapter. De- 
posits for many of these minerals are so ubiquitous that it is possible to ex- 
ploit those with the most favorable access to markets. This is particularly true 
of minerals in the stone-clay-sand-gravel category and these make up almost 
two-thirds of the value of the output of all non-metallic minerals. Two-thirds 
of the output of these sand and gravel type minerals goes to construction and 
final demand. As a consequence, the dominant change in access explaining 
the shifts in non-metallic mining is the westward shift in the market. This 
shift has been revealed for population but is even more marked for construc- 
tion. The change in regional access to markets seems to have prompted a more 
intensive exploitation of the resource deposits near the expanding market areas. 

The changes in population and associated markets, however, cannot explain 
the direction and dimension of all of these shifts. Individual minerals with 
relatively concentrated deposits are often an important element in the explana- 

29 When we try to explain the shifts in non-metallic mineral mining, however, we run into 
the same complicating element presented by the analysis of metallic minerals. There are 
some 60 different minerals which are brought together in this group. This lack of homo- 
geneity makes a thorough analysis far beyond the scope of this study. 

30 This high correlation may be deceiving to a degree because of the fact that non-metallic 
minerals is a family group made up of so many different minerals. Some of these have 
highly concentrated markets as well as highly concentrated inputs (like potash and phos- 
phate). However, the fact that there are so many of these activities has the result that they 
tend to average out spatially and strengthen the correlation rather than weaken it. 



332 / Regional Distribution of Economic Activities, 1939-1954 




|Hy|| Absolute upward shift 

2] Relative downward shift 
; I Absolute downward shift 



Figure 56. Net Shift in Non-Metallic (other than fuels) Mining Employment, 
1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 104.4. Total net shift as % of 1939 non-metallic mining 
employment: 19.1. Source: Appendix Table I. 



tion. While California records a large net upward shift (13.5%) in part be- 
cause of sizeable increases in the construction minerals, other factors are (1) 
that it has major deposits of borates, which is a "rapid-growth" mineral, and 
(2) that it has expanded its share of the output of both salt and gypsum. The 
large increase in New Mexico arises primarily from the fact that it supplies 
90% of the nation's potash output, and potash has been a growth mineral. 
Further, Texas has increased its salt output four times as much as the rest of 
the nation. Arkansas shows up strong because of an important discovery of 
a new bauxite deposit which has made it the leading producer of this mineral. 
Florida's performance stems in part from the fact that it produces three-fourths 
of the nation's phosphate — still another growth mineral. We cannot attempt 
to evaluate the changes leading to these shifts. 



REGIONAL IMPACT OF MINING ACTIVITY 

The general regional influences reviewed here have touched a large number 
of states that have some mining activity. It should be recognized, however, 
that in only a few of these states have the shifts in mining employment been 
sufficiently important to make an impact upon total employment. There were 
only eleven states (Table 134) in which the shifts accounted for more than 



Regional Distribution of Mining Activity / 333 

5% of the total differential effect; in all but three of these states the shifts in 
mining employment took the same direction as the shifts in total employment. 31 

Most striking is the fact that the performance of mining employment ex- 
plains the behavior of one of the major growth regions and one of the major 
relative decline regions. 

Of the states which have shown the greatest growth in total employment, the 
Southwest in general and Texas and Louisiana in particular seem to have 
owed a major part of their growth to the mineral sector of the economy. Texas 
and Louisiana, which had two-thirds of the net upward shift in mining em- 
ployment, were the only states where the share of mining and employment in- 
creased significantly more than the share* of population. This would suggest 
the possibility of mineral resources being an important lead factor in the growth 
of these areas. Oil and gas and their geological associates — sulphur and salt — • 
were chiefly responsible for this performance. 

The Appalachian states comprised one of the major areas of relative decline, 
with a marked downward shift in coal-mining employment. Pennsylvania, West 
Virginia, and Kentucky account for about 75% of the net downward shift in 
total mining employment. The proportionality effect was dominant here. As 
a matter of fact, Kentucky and West Virginia strengthened their relative posi- 
tion in coal-mining and displayed a net upward differential shift for that in- 
dustry. However, this effect was slight compared to the large losses associated 
with the proportionality effect. There was also a downward differential shift 
in petroleum and natural gas and, though less important, in non-metallic min- 
erals in these states. 

A review of the regional influence of mining activity based upon the ma- 
terials in this chapter cannot provide a complete picture, however. California 
is a case in point. It exhibited the greatest over-all growth in the nation for 
the 1939-1954 period but showed only a modest relative upward shift in mining 
employment. This shift was based almost entirely upon the proportionality 
effect; the only differential upward shift was in non-metallic minerals. This does 
not mean, however, that mineral activities have not made important contribu- 
tions to the growth of California. California shows an amazing diversity of 
output in both metallic and non-metallic minerals and produces about 10% of 
the total oil and gas output for the nation. This diversified resource base has 
undoubtedly contributed to the state's economic growth. Furthermore, the shift 
picture can be deceptive. For example, the relative net downward shifts recorded 
for petroleum and natural gas are still consistent with sizable absolute in- 
creases. If petroleum and natural gas have any important multiplier relation- 
ships with other activities, these increases could support growth that would be 
reflected in other dimensions. 

If we could examine these cumulative factors more adequately there are a 
number of questions concerning the regional influence of mining activity for 
which we might wish to seek answers. (1) Does the considerable importance 

sl The upward shifts in mining employment displayed by Louisiana, Minnesota, and Wyo- 
ming ran counter to, and were overshadowed by, the general shifts in other employment 
sectors in those states. 



334 / Regional Distribution of Economic Activities, 1939-1954 

of substitute sources and substitute products plus the possibility of exhaustion 
place regions which are mining specialists in a more vulnerable position than 
other regions? Is it possible that their growth behavior tends to be more 
erratic and less predictable than other types of regions? (2) Since mining 
regions are usually subject to important external as well as internal stimuli, 
does this have any special significance for the growth performance of mining 
regions? (3) Since mineral inputs often exhibit significant weight-loss when 
processed and since most of these inputs go to intermediate processing activities 
rather than final demand, do mineral specializing regions tend to have a higher 
multiplier effect working for them than some others (like agricultural regions, 
for example) ? (As we proceed to an examination of the components of manu- 
facturing activity in Texas we may gather some evidence of this.) (4) There 
seems to be a trend underway that displays an expanding number of substitu- 
tion possibilities in the minerals industries. Discovery and technology are 
enlarging the range of mineral product substitutes. Technology often has the 
effect of making feasible alternative mineral sources that were previously non- 
economic. Does this mean that the mold for economic activity imposed by the 
distribution of natural mineral resources is becoming more flexible? Is the 
spatial influence of these resource sectors becoming less important, and what 
does this imply for the regional dimensions of future growth? 
Topics such as these require volumes rather than chapters. 



21 / Regional distribution of 
agricultural activity: 

(1) FARMING IN GENERAL 



In terms of employment numbers, agriculture was the largest of the ten broad 
industry components of total employment in 1939; and in 1954 it was exceeded 
only by the manufacturing sector. 1 But, as in mining both the absolute and the 
relative importance of agriculture in the nation's total employment is declining. 
In this chapter we shall consider the factors that are operative in this develop- 
ment and the regional consequences of it. 

3 An examination of unpublished long-term resource output and employment indices 
compiled by Neal Potter and Francis T. Christy, Jr., at Resources for the Future, Inc., 
reveals that the years 1939 and 1954 were reasonably "normal" for purposes of comparison, 
with respect to output and employment. 









A 




B 




Sector 






1939 




O 1954 


C 




Av 


1937-41 


Av O 1952-56 


(B-A) 





= Output 


in physical terms, 


selected 


[ major components 




All agriculture 






97.6 




98.5 


.9 


All crops 






94.3 




97.5 


3.2 


Wheat 






88.4 




91.0 


2.6 


Corn 






100.2 




94.1 


- 6.1 


Cotton 






89.5 




93.4 


3.9 


Livestock 














Dairy products 






99.3 




101.3 


2.0 


Hogs 






108.8 




97.3 


-11.5 


Cattle 






100.3 




101.6 


1.3 



As indicated in Columns A and B of the table above, output in major agricultural sectors 
tended to be somewhat lower for the years 1939 and 1954 than for the average of neighbor- 
ing years. 

Column C represents the difference between the ratios in Columns A and B. If there is 
little difference in the ratios, one can assume that conclusions resulting from comparison 
of the 1939 and 1954 figures are not substantially impaired by the fact that these years were 
somewhat atypical. Where the difference is positive, the conclusions may be overstated with 

/ 335 



336 / Regional Distribution of Economic Activities, 1939-1954 

Regional Influence of Agriculture Reflected in 
the Proportionality Effect 

Between 1939 and 1954 thirty-one states experienced a net downward propor- 
tionality shift in total employment as a consequence of their specialization in 
sectors of economic activity that have grown more slowly than others or have 
actually declined for the nation as a whole over these years (Figure 47) . Mining 
specialization, as we have seen, contributed to this downward shift in about half 
of these states, but it was the major cause of the proportionality effect in only 
two of them — Nevada and Utah. In twenty-nine of the thirty-one states the net 
downward proportionality shift was attributable primarily to specialization in 
agriculture. Table 139 shows agriculture's share of total employment in each 
of these states in 1939 and in 1954. 

The agricultural labor force of the nation declined by 14.7% over these years. 
But the decline in agricultural employment does not have its origin in the cur- 
rent period; it goes back over several generations. Since 1870 the share of 
agriculture in the nation's total labor force has fallen off by more than two- 
thirds,, and its share in total income by more than a half. 

Several long-term trends in supply and demand help to explain these devel- 
opments. On the supply side, the major trends may be summarized as follows: 

(1) There has been a pronounced long-run trend for agriculture to cover 
fewer economic activities than formerly. This has been a function largely of 
technological advances and the increased specialization that characterizes an 
advancing society. To a considerable extent, activities that once belonged to 
agriculture have been taken over by the manufacturing and the trade and service 
sectors of the economy. The relative importance of agricultural employment has 
declined significantly as a result. 2 

(2) The value-added factor also helps to explain the relative decline of agri- 
culture. The more extended fabrication that agricultural products undergo 

respect to the period — where negative, understated. Thus, for all agriculture, both years 
are slightly subnormal and the slight degree of difference suggests that comparative con- 
clusions for the period are not seriously compromised. Major departures are apparent with 
respect to corn and hog output, but the differences being negative, comparisons involving 
these two products during the period 1939-1954 would tend to be understated. 

Potter-Christy agricultural employment data tend to support the assumption of normalcy 
for these years: 





A 

E 1939 


B 

E 1954 


C 
(B-A) 




E1937-1941 


E1952-1956 




All Agriculture 

Employment 

Manhours 


100.7 
100.4 


97.8 
99.3 


- 2.9 

- 1.1 



2 Cf. John H. Davis and R. A. Goldberg, A Concept of Agribusiness (Boston: Harvard 
University Press, 1957). Ronald Mighell has estimated that employment in agriculture and 
its associated pre-farm and post-farm activities amounted to about twice that of agriculture 
alone in 1950. See Ronald L. Mighell, American Agriculture: Its Structure and Place in 
the Economy, A Volume in the Census Monograph Series, (New York: John Wiley and 
Sons, 1955) , p. 11. 



Regional Distribution of Agricultural Activity (1) / 337 

Table 139. Agricultural employment as percentage of total employment, 1939 
and 1954, for states with net downward proportionality shifts in total employ- 
ment, 1939-1954, because of specialization in agriculture 



State 



1939 

% 



1954 

% 



United States Average 

Mississippi 
Arkansas 
North Dakota 
South Dakota 
South Carolina 
Alabama 
Georgia 
Idaho 
Kentucky- 
North Carolina 
Texas 
Oklahoma 
Louisiana 
Tennessee 
Nebraska 
New Mexico 
Iowa 
Kansas 
Montana 
Minnesota 
Wyoming 
Virginia 
Vermont 
Arizona 
Wisconsin 
Missouri 
Oregon 
Colorado 
West Virginia 



27.5 



74.3 
70.1 
68.3 
57.6 
56.3 
54.6 
50.4 
49.1 
48.8 
46.9 
46.9 
46.8 
46.5 
46.2 
46.1 
46.0 
45.6 
42,8 
42.4 
40.0 
36.0 
35.9 
35.1 
34.8 
34.5 
32.7 
31.4 
30.0 
28.1 



16.9 



59.2 
54.9 
55.4 
50.5 
31.8 
32.1 
23.4 
39.5 
34.5 
34.8 
25.4 
31.2 
29.6 
33.2 
35.1 
27.5 
35.9 
27.7 
29.5 
28.4 
23.2 
22.2 
24.5 
22.7 
23.6 
21.8 
19.8 
19.5 
17.6 



Source: Appendix Tables Fl and F2. 



before reaching the final consumer generally adds to aggregate output without a 
commensurate increase in agricultural output. 

(3) Great increases in worker productivity have helped bring about a relative 
decline in the importance of agricultural employment. 3 There have been enor- 
mous increases in efficiency through mechanization and other technological 
improvements. Agricultural land and labor could be made to go further for 

This increase in worker productivity, however, would not explain the relative decline in 
agricultural employment if all sectors of economic activity improved worker productivity 
by an equal amount. Manufacturing has similarly improved worker productivity. The 
services and trades, however, have lagged behind agriculture in this respect. Since 
agriculture has experienced a more rapid technological development than the services and 
trades, we expect this sector to lose labor over time to the services and trade sectors. 



338 / Regional Distribution of Economic Activities, 1939-1954 




Regional Distribution of Agricultural Activity (1) / 339 

another reason: tractors don't eat hay. Between 1918 and 1953, substitution of 
tractor power for horse and mule power released about 70 million acres of land 
formerly used to produce feed for draft animals. The release of agricultural 
labor was similarly impressive. 

Long-term trends operating on the demand side of the market have also con- 
tributed to the relative decline of agricultural employment. Three of these are 
worth special mention. 

(1) Consumption of farm products as a whole is not very responsive to 
changes in income; income elasticities are relatively small. This is seen most 
readily in the case of food products. Expenditures for food at retail stores and 
restaurants have been increasing at a rate which is more or less in proportion 
to increases in income. But these expenditures include many services of pro- 
cessing and distribution. And it is these services that have been increasing with 
rising per capita income rather than the consumption of farm products. The 
quantity of food consumed per person increases very little as incomes rise. 
Agriculture's share of total demand declines progressively as incomes and total 
demand grow. 

Recent studies of per capita expenditures for food in relation to income (both 
in real terms) covering the periods 1929-41 and 1948-56 suggest an income 
elasticity of demand for food of .44; a 10% increase in real per capita income 
has resulted in an average increase of around 4%% in food expenditure per 
person. However, this total is made up of two parts ; the bill for processing and 
marketing and the farm share (with some two-thirds of the total currently going 
to processing and marketing) . For the dates studied, the income elasticity for 
the service part was found to be .775 (the "quantity" of services increased by 
about 7%% with a 10% increase in income), while the income elasticity of 
demand for food at the farm level was less than .15 (a 10% increase in real 
consumer income per person resulted in a gain of only 1%% in per capita 
expenditures for food). The very low income elasticity of demand for farm 
products at the farm level has resulted in a long-run decline in the farmers' share 
of total expenditures, and even of expenditures for farm products as such. 4 

Not only has the income elasticity of demand for farm products been low in 
recent decades, but it has been falling. Thus data for market value of all food 
in relation to disposable money income suggest the following income elasticities 
of demand: 5 

1929-41 1948-57 Spring 1955 (cross section 

.68 .38 ^25 data) 

4 Rex F. Daly, "Demand for Farm Products at Retail and the Farm Level: Some Empirical 
Measurements and Related Problems," Journal of the American Statistical Association, Vol. 
53 (September 1958) pp. 665-68; and "The Long-Run Demand for Farm Products," Agri- 
cultural Economics Research, Vol. 8 (July 1956), pp. 73-91. 

s Marguerite C. Burk, "Some Analyses of Income-Food Relationships," Journal of the 
American Statistical Association, Vol. 53 (December 1958), pp. 905-27. 

Figure 57 — opposite. Major Types of Farming in the United States. 

Source: U. S. Department of Agriculture, Agricultural Research Service. 



340 / Regional Distribution of Economic Activities, 1939-1954 

While estimates of this type involve all sorts of statistical problems, the general 
magnitudes are highly suggestive and help to explain developments in the farm 
sector of the economy. 6 

There is an obvious physiological limit to per capita consumption of food, 
whether considered in terms of weight, calories, or other constituents. The popu- 
lation of the United States has probably come closer to reaching this limit than 
any other group in history, although it must be recognized that many Americans 
lack an adequate diet. Also, of course, changes in eating habits influence food 
purchases. 

(2) There has been a long-run decline in agriculture's share of American 
exports. This can be traced to a number of things such as the tariff policy of 
the United States, 7 agricultural protection abroad and modifications in com- 
parative advantage and the terms of trade to favor manufactured products. This 
declining trend has been interrupted occasionally by abnormal war demands but 
has been a fairly consistent long-run element in agricultural demand over a 
number of decades. 

(3) Lastly, for a number of agricultural products, there has been sharp com- 
petition from synthetic and natural substitutes of non-agricultural origin. The 
case of the synthetic fibers is a good example. 

In short, significant long-term trends in supply and demand have reduced the 
relative importance of the agricultural sector of the economy. As a conse- 
quence, the regions that have depended largely on agriculture have tended to 
lose out relative to regions with better access to manufacturing markets and 
basic intermediate inputs. 8 In spite of this generally depressing effect of agri- 
cultural specialization, some regions received significant assistance from agricul- 
ture in generating net upward shifts in total employment. This is so because 
there were significant interregional differential shifts in agricultural activity. 

6 See also James P. Cavin, "Projections in Agriculture," Long-Range Economic Projection. 
Studies in Income and Wealth, Vol. 16 (Princeton: Princeton University Press, 1954), pp. 
107-30; U.S. Dept. of Agriculture, Food Consumption of Households in the United States; 
Northeast; North Central Region; South; and West. Household Food Consumption Survey 
1955 Reports 1 to 5, (Washington: U.S. Dept. of Agriculture, 1956). University of Penn- 
sylvania, Study of Consumer Expenditures, Incomes and Savings, 1956. (1950 food data in 
volumes III and XII, Spring 1951 food data in Vol. XII.) 

7 Since the first half of the 1939-54 period was dominated by World War II and the 
succeeding international adjustments, agricultural products constituted an abnormally low 
proportion of total exports (an eight year average of 20%). When compared to the latter 
half of the period (an eight year average of 25%), it is apparent that during this period 
agricultural products constituted a rising proportion of total exports. 

However, the long run trend is one of declining proportion. During the five years follow- 
ing World War I (1919-23) the average proportion of agricultural exports was 46%, while 
during the similar period following World War II (1946-50) the proportion was 28%. U.S. 
Department of Agriculture, Agricultural Statistics, 1957 (Washington: Government Printing 
Office, 1958), Table 808, p. 702. 

8 It should be noted that the multi-state regions used for analytical purposes in this study 
do not always correspond well with the generalized agricultural regions (See Figure 57). If 
one were focusing essentially on the major features of the agricultural sector of the national 
economy and seeking to understand the factors that have given shape to the economic 
geography of American agriculture, this would be a serious shortcoming. The purpose of 



Regional Distribution of Agricultural Activity (1) / 341 




Absolute upward shift 
\f//\ Relative upward shift 

Absolute downward shift 



Figure 58. Net Shift in Agricultural Employment, 1939-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 34.7. Total net shift as % of 1939 agricultural 
employment: 5.1. Source: Appendix Table H. 



Regional Influence of Agriculture Reflected in the Differential Effect 

Figure 58 shows the net shifts in agricultural employment, by states, between 
1939 and 1954. 9 It is clear, when this figure is compared with Figure 46, that 



this study is somewhat different, however; it seeks to understand differential regional 
development in terms of the array of economic activities and their changes. Agriculture is 
but a part of this total picture, and our concern is with its role in the total pattern of 
regional economic development. Thus, it is this kind of "vertical" analysis that requires that 
the geographical regions be held constant throughout. 

For a finer "grain" of analysis, we have focused on shift phenomena at the state level. A 
glance at the maps portraying the nature and direction of the shifts in activity allows the 
reader to aggregate by states visually in any manner of interest to him. Thus, in Figure 58 
one could conclude that the Corn Belt states generally enjoyed a modest net upward shift in 
agricultural employment during the period 1939-54, while the Cotton Belt states suffered a 
modest-to-extensive net downward shift during the same period. 

9 In evaluating the figures that are presented on agriculture, it should be kept in mind 
that agricultural employment differs from most other employment. In most activities, persons 
are considered employed only when actually working for wages or directing such workers. 
In agriculture, many "workers" are essentially persons not working elsewhere. Farming 
conceals much unemployment and underemployment and their extent is not constant from 
period to period. Thus agricultural "employment" data must be viewed with this considera- 
tion in mind. 



342 / Regional Distribution of Economic Activities, 1939-1954 

agriculture has played a significant part in the differential net shifts in total 
employment. Of the twenty-six states with net upward differential shifts in 
total employment, nineteen show relative gains in agriculture; these nineteen 
states account for 77% of the net upward shift in agricultural employment. The 
correlation is not so high for the twenty-two states that show net downward 
shifts in total employment. However, in over half of these states, agriculture 
contributed to the relative decline; 83% of the net downward shift in agricul- 
tural employment occurred in states which also show a net downward differen- 
tial shift in total employment. 

Additional evidence supports the impression that agricultural employment 
shifts have been an important part of the observed total differential shifts. 
Table 140 identifies twenty-six states in each of which the agricultural shift 
accounted for more than 10% of the total shift (computed without regard to 



Table 140. Differential shifts in agricultural employment as percentage of 
total employment for states with largest net differential shifts in agricultural 
employment, 1939-1954 



State 



Agricultural shift in state as % of 
total shift (without regard to sign) 
of 1— digit employment categories. 



South Carolina 

Georgia 

Alabama 

Idaho 

South Dakota 

Tennessee 

Mississippi 

Nebraska 

North Carolina 

Iowa 

Utah 

West Virginia 

Arkansas 

Indiana 

Arizona 

Wisconsin 

Colorado 

Oregon 

Maine 

Minnesota 

Montana 

Kansas 

New Mexico 

North Dakota 

California 

Texas 



70.1% 

64.3 

56.8 

54.6 

49.1 

45.3 

36.2 

35.9 

30.1 

29.6 

28.0 

26.7 

25.3 

22.9 

22.0 

21.9 

20.9 

20.7 

19.9 

17.3 

16.1 

14.3 

14.1 

12.6 

12.3 

11.7 



Source; Table 157. 



Regional Distribution of Agricultural Activity (1) / 343 

sign) of the 1 -digit employment components for that state. The states most 
significantly affected by the performance of agriculture are concentrated in 
the Southeast, Plains, and Mountain regions. 

There are no usable figures for the 1939-54 period on employment in the 
component segments of agriculture, such as we used in our examination of 
mining employment. In order to present a more detailed analysis, therefore, 
we have used data on the value of agricultural products sold in each of the main 
agricultural categories. These data are presented in Table 141. The net shifts 
in value, or the over-all departure of states from the national norm, have been 
significant in each of the main categories. Departures from the national average 
change have been greatest in vegetables and poultry and products and smallest 
in dairy products. 

The state-by-state shifts in the value of all farm products sold are shown in 
Figure 59. Like the data on agricultural employment, these data support the 
general picture of a westward shift in agriculture. The states in the East on 
the whole display downward shifts, and those in the West predominately upward 
shifts. Further, 86% of the net upward shift in the value of farm products 
sold took place in states that also displayed upward shifts in agricultural 
employment. 

There are, however, marked discrepancies between the shifts in the value 
of farm products sold and the shifts in agricultural employment. Several Moun- 
tain, Plains, and Great Lakes states that show net upward employment shifts 
show net downward shifts in value of farm products sold. Just the reverse is 




^] Absolute upward shift 
-.■•'.{ Relative downward shift 



Figure 59. Net Shift in Value of All Agricultural Products Sold, 1940-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change; 9.2. Total net shift as % of 1940 value of all 
agricultural products sold: 24.7. Source: Appendix Table J. 



344 / Regional Distribution of Economic Activities, 1939-1954 

true for some of the Southeastern states. The most marked contrast is the sharp 
upward shift in agricultural employment in Tennessee and North Carolina in 
the face of relative downward shifts in the value of farm products sold. 

There can be many reasons for these discrepancies. (1) Some labor-intensive 
agricultural commodities may have made relatively limited gains in agricultural 
values over the period. (This seems to be the case for dairy products, which 
may help explain the discrepancy in the Great Lakes states.) (2) There are 
substantial regional differencies in part-time farming and opportunities for off- 
farm employment that might serve to explain some of these differences. (3) 
Mechanization has proceeded more rapidly during this period for some agricul- 
tural products than for others. (4) Interregional differences in farm tenancy 
and other institutional and cultural patterns may have some influence. (5) 
Historical factors arising out of farm adjustments in previous periods and the 
differences in timing of the same adjustment in different regions may play an 
important role. (6) A whole host of "accidental" factors that are more or less 
peculiar to a specific region may exert their influence. 

We cannot undertake here the complex job of translating value phenomena 
into employment phenomena. 10 We proceed, however, to examine in greater 
detail the shifts displayed by the value of farm products sold for two reasons. 
First, there is no better alternative for providing disaggregated detail within 
the practical time limits with which we are faced. Second, the within-component 
changes reveal many things of direct value in arriving at an understanding of 
the shifts in agricultural employment. 



10 To highlight the types of problems involved, let us look briefly at the most intriguing of 
these discrepancies, displayed by North Carolina and Tennessee. The reader should keep 
in mind that the base for the shift is the average performance for the nation and this net 
upward shift is primarily evidence of a less rapid decline in farm employment in these 
states. (Even Tennessee's evidence of an absolute upward shift is unimportant. Less than 
4% of the net upward shift for this state was generated by an absolute increase.) With 
this in mind we should recognize that North Carolina and Tennessee made an earlier ad- 
justment in their farm economy than the Deep South. Thus, in the current period their 
decline in agricultural employment was less marked and became recorded as a net upward 
shift. We have examined several sets of statistical data that support this conclusion. Further, 
Tennessee and North Carolina have had fewer farm tenants than the Deep South, and farm 
tenants tend to leave the soil more readily than farm owners. Secondly, there is some 
evidence that the set-back in manufacturing employment that occurred in the Manufacturing 
Belt in 1954 may have sent a number of transient industrial workers back to their hill 
farms in time to be counted in the census. Third, Tennessee and North Carolina are the 
most industrialized of the Southeastern states and present more off-farm employment oppor- 
tunities that encourage part-time farming. This tends to keep up agricultural employment 
without equal support to farm production. Fourth, dairy and poultry are reasonably con- 
centrated activities in these states and they tend to be labor intensive although they have 
not made the gains in value that other agricultural products have. Fifth, much bottom land 
in these two states was in production at the beginning of our period that was inundated 
and out of production by 1954. Much of this production was in high value vegetable crops. 
This could help explain the relatively depressed character of value of farm products relative 
to farm employment. 

One can see that a wide variety of factors can serve to explain the discrepancy in 
each case. 



Regional Distribution of Agricultural Activity (1) / 345 



1^-3 

O eft O 

o 



o w o 

o 



T3 



o 

s 

o 



C r- "5* O CO 

p-h 72 ri, 23 rt 

« ^ i •? s 



bC eu 



— «)■<# 



CO 



0) 

u 

c 

■K 

o 

ft 

s 






3 Sb lo .2 £ 

to JS Tf '2 S 

-q -c on £ o 

<J W H s_ -O 









^cNfgiofg^c0COH\o 

*—< HlNHrtHIMHCO 



OJ "* <N* vd CO © <tf 00* CO 
t^VOC^r-Hr-H <N CO t"~ 

+i +1 +1 +1 +1 +1 +1 +| +| 



cqvqcococqcsjHh-H 

^ 00 rH CO* LO © CO LO to CO 

~^0(Noma^Nco 



0>— i O M ^Ji \o H N p- 1 

^\om^OMH\o\oa 



ocovoonoolocovolo 



"^ O © 



tM tJ< i— i co t- 



£ a 
« .2 



o rs 
Q £ 



U 



fO \D * \0 h rf «j (^ O h 
*"*' rS** ©\ i-H to h h 

CS 



OOOt^onI^cOi— I VO 

pJoncc^' Hoodoo 

C O CO r-1 CM 



MHOLoaoow^a 

CONOOiMHlOhcO 

VO^MMHHiOCO 



to 



to 
3 "" 



1:3 o 



"^ b 



: o 1 1 | 



'k o -Q 



„ O 

O •« 



ft 

Ph 
<i 



346 / Regional Distribution of Economic Activities, 1939-1954 

Table 142. Percentage change in value of agricultural products sold for 2-digit 
agricultural sectors, 1940-1954 



Fruits and nuts 305.8% 

Field crops 301.6 

Livestock and products 275.7 

All Agricultural Products 268.8 

Horticultural specialties 250.8 

Poultry and products 245.1 

Farm forest products 233.1 

Vegetables 223.3 

Dairy products 198.2 

Source: U. S. Census of Agriculture 1940, Vol. Ill, T. 8, pp. 905-11; U. S. Census of 
Agriculture 1954, Vol. II, General Report, Chap. 9, "Value of Farm Products." 



1 -Digit Proportionality and Differential Effects in the Value of 
Farm Products Sold Based Upon 2-Digit Components 

In Figure 59 we have shown the net shift in value of all agricultural products 
sold. In order to introduce further detail into our picture of structure and 
change in American agriculture, we shall now subdivide this total value into 
its eight major components — field crops, fruits and nuts, horticultural speciali- 
ties, dairy products, poultry and products, livestock products, and farm forest 
products. By computing a shift pattern for each of these components and 
summing them for each state, we arrive at the total differential net shift in the 
value of farm products sold. Since this figure does not typically exhaust the 
total shift portrayed in Figure 59, the balance represents the proportionality 
effect. The proportionality shift is presented in Figure 60 and the differential 
shift in Figure 61. 

The proportionality effect contributes only about 20% of the total shift in 
the value of farm products sold. However, it assumes significant proportions 
in specific regions. The net downward proportionality shift is concentrated in 
the Northeast — the New England, Middle Atlantic, and Great Lakes states. Two 
states, Wisconsin and New York, account for 43% of it, and Pennsylvania 
for 12%. 

Table 142 points out the reasons for this effect: dairy products, poultry prod- 
ucts, vegetables, horticultural specialities, and farm forest products are the 
principal slow-growth sectors of agriculture when measured by the value of 
agricultural products sold. Table 143 shows that these are also the sectors with 
the greatest association with markets. It is no surprise then, to find the net 
downward proportionality shift concentrated in the highly populous market 
centers of the Northeast. Since dairy products create well over half of the total 
value of these slow-growth commodities, they tend to dominate the net down- 
ward proportionality shift. It is also no surprise to see the concentration in 



Regional Distribution of Agricultural Activity (1) / 347 




Upward sh 
| J Downward shift 



Figure 60. Total Proportionality Net Shift in Value of Agricultural 
Products Sold, 1940-54. 




Upward shift 
| | Downword shift 



Figure 61. Total Differential Net Shift in Value of Agricultural 
Products Sold, 1940-54. 



348 / Regional Distribution of Economic Activities, 1939-1954 

Wisconsin and New York. These are the nation's dominant dairy states. The 
largest net upward proportionality shifts are seen to have occurred in North 
Carolina, Texas, Iowa, Mississippi, and Arkansas; in all of these states there 
is specialization in the rapid-growth sectors. 

An understanding of this proportionality effect requires some knowledge of 
the reasons why some sectors of agriculture expand more than others. The 
modifications of total demand and supply factors that bring about this result 
are legion. 



Table 143. Rank correlations of each state's share of selected agricultural 

activities with each state's share of population, 1939 and 1954 

Agricultural activity 1939 1954 

Total value of agricultural products sold .666 

Total agricultural employment .708 .649 
Value of agricultural products sold in* 

Horticultural specialities .816 .859 

Dairy products .736 .762 

Poultry products .754 .751 

Vegetables .759 .694 

Fruits and nuts .723 .641 

Field crops .544 .383 

Other livestock products .368 .324 

* Figures for 2-digit components of "Value of Agricultural Products Sold" are for 1940. 
Source: See Sources for Appendix Tables Fl, F2, H and J. 



One factor is, again, income elasticity of demand. 11 It is true that for agri- 
cultural products as a whole, the income elasticity of demand is less than one. 
However, there can be significant variations among agricultural products. Fruits 
and nuts and livestock products tend to show relatively high income elasticities. 
This may be an important factor in explaining the growth performance of these 
sectors. On the other hand, vegetables and dairy products also tend to have a 
higher income elasticity than field crops, so some other factors must play a role. 

One aspect of the observed rates of growth may be explained by an important 
substitution. For some uses, vegetable oils have proved to be a potent substitute 
for dairy products. During our period the repeal of restrictive legislation on 
oleomargarine and the stimulus provided by the war undoubtedly created a 
significant adjustment. Since vegetable oils are classed under field crops, this 
development could be an important factor in explaining the relatively poor 
performance of dairy products vis-a-vis field crops. 

The large increases in farm productivity that have accompanied mechaniza- 
tion may also be a factor. In general, field crops have been far more suscepti- 

n See Theodore W. Schultz, The Economic Organization of Agriculture, (New York: Mc- 
Graw-Hill Book Co., 1953), pp. 175-94. 



Regional Distribution of Agricultural Activity (1) / 349 

ble to mechanization than animal livestock activities. Dairying in particular 
has shared less in the productivity increases that have accompanied farm me- 
chanization. We also know that the farm price-support program has not been 
applied uniformly to all agricultural commodities. For example, storable com- 
modities are easier to subsidize than non-storable commodities. Perhaps field 
crops were nudged into the rapid-growth category by the role of farm price- 
supports. It is also possible that short-run distortions in farm output were 
caused by the changes in agricultural demand that accompanied the war. 

When we turn our attention to the differential net shift, we can see clear 
evidence of the point already made — the differential effect dominates the net 
shift in the value of farm products sold. The configurations in Figure 61 match 
very closely those in Figure 59. Only Texas, Colorado, and North Dakota 
experienced upward shifts that were created because of their specialization in 
rapid-growth products alone. It can also be easily established that the states 
showing the greatest upward and downward shifts in the value of farm products 
sold also show the greatest upward and downward differential shifts. The 
explanation of these shifts must lie in the differential regional behavior of the 
components of agricultural activity. 



Access Characteristics of Agricultural Activities 

As a resource activity, agriculture must be located at the site of the immobile 
resource attributes it exploits. The range of substitution possibilities in culti- 
vation is so great, however, that these immobile attributes exercise only moder- 
ate influence upon the specific location of agricultural activity. Other factors 
can exert considerable influence. There are several reasons why this is so. 

(1) The resource that agriculture exploits is land — land and all of its attri- 
butes of fertility, topography, water-holding capacity, and its associated pre- 
cipitation and temperature patterns. However, arable land in the United States 
is nearly ubiquitous from a regional point of view. True, arable land is not 
all of the same quality. When we consider specific agricultural functions and/or 
small areas, the amount of land specifically suitable may be much less ubiqui- 
tous. The general picture, however, is one of a wide range of substitute input 
sources. 

(2) The fact that labor and capital can be substituted for land resources to 
a considerable degree enlarges the range of substitution possibilities. The 
appropriate temperature conditions can be provided by hot houses, the moisture 
conditions by irrigation, and the fertility conditions by chemical fertilizers in 
areas where the natural attributes are found wanting. Such factors as these 
can open the way for a larger influence upon location by non-resource attributes. 

(3) Another attribute of agriculture is that it is based on a renewable re- 
source. It is true that large land areas have been depleted of much of their 
productive capacity and that important regional shifts have been associated 
with this depletion. In the main, however, advantage is taken of the replen- 
ishible characteristic of agricultural land with two consequences of significance 



350 / Regional Distribution of Economic Activities, 1939-1954 

for regional analysis. First, regions that specialize in agriculture need not be 
in as serious danger as mining areas of economic stagnation due to resource 
exhaustion. Second, there is a choice between rational husbandry and an 
essentially "mining" type of activity. The former, which is increasingly the 
choice, means that purchased inputs become more important in agriculture 
than in other resource activities. Thus, the range of substitution possibilities 
on the supply side are greater, and this is another factor contributing to greater 
locational flexibility for agriculture. 

(4) More than almost any other activity, agriculture is subject to important 
external substitution alternatives. By its very nature, it cannot be as concen- 
trated an activity as mining or manufacturing. This extensive use of land 
surface in agriculture makes this activity more liable than most to the influence 
of "rivalry" costs. All agricultural activities cannot locate at the same point. 
Competition for the most central sites bids up land rents. Producers of certain 
agricultural products need a central location more than others and will be able 
to pay a higher rent. These functions appropriate the most favorable locations 
and, by imposing upon them an undesirable external cost element, displace 
other activities to peripheral areas. (This relationship applies not only to 
alternative agricultural uses but to alternative urban household and industrial 
uses as well.) Thus, the location of agricultural production is modified in its 
detail by these important competitive relationships influencing the relative 
availability of land inputs on the supply side. An understanding of the regional 
access characteristics of an agricultural activity, therefore, involves considerable 
understanding of the characteristics of rival agricultural systems. 

Certain of the special characteristics of agriculture may be seen by comparing 
the inputs into agriculture with the inputs of mining, the other major resource 
sector. Such a comparison is made in Table 144. The mining inputs support 
an almost classical view of a resource activity. Some 85% of the purchased 
inputs come from the final demand sector, which is primarily labor. 12 Another 
15% is split between manufacturing and service. This is not far removed from 
the simple picture of labor being applied to natural resources to create value 
— the traditional view of resource exploitation. In agriculture, the contrast is 
marked. Labor inputs are significant, but they account for less than half of 
the total. 13 Significant purchases are made from manufacturing, the services 
(especially transportation), and agriculture itself. 

Consider these inputs further. Purchases from the manufacturing and service 
sectors account for half of the non-labor farm expenses. As short a time as 
thirty years ago, agriculture purchased only from 20% to 30% of its non-labor 
farm inputs. 

Less than a century ago the farmer not only produced food and fiber but 
also created most of his own production supplies, provided his own storage, and 

12 The final demand sector includes tax payments to governments as well as payments to 
labor. Labor payments in this case are just under 80 per cent. 

"The data on this point in the BLS table represent rough estimates at best. There are 
serious difficulties involved in measuring the labor inputs of self-employed farmers and 
partially employed workers. 



Regional Distribution of Agricultural Activity (1) / 351 

Table 144. Percentage distribution of inputs and outputs for major industry 
groups, 1947 



a. Distribution of Inputs 



Output 










Services 






Agricul- 




Manufac- 


and 


Final 


Input source 


ture 


Mining 


turing 


trade 


demand 


Agriculture 


25.7 


0.0 


9.9 


.8 


3.1 


Mining 


0.1 


1.0 


3.4 


1.1 


0.3 


Manufacturing 


10.2 


8.7 


39.4 


15.3 


27.9 


Services and trade 


14.6 


6.2 


8.8 


19.9 


44.4 


Final demand (mainly labor) 


49.3 


85.8 


38.2 


63.0 


24.6 



Total 



100.0 



100.0 



100.0 



100.0 



100.0 



b. Distribution of Outputs 



Industry 











Services 




Output 


Agricul- 




Manufac- 


and 


Final 


Distribution 


ture 


Mining 


turing 


trade 


demand 


Agriculture 


25.7 


.5 


2.1 


3.0 


6.7 


Mining 


0.0 


1.0 


0.4 


0.3 


2.6 


Manufacturing 


47.9 


73.4 


39.4 


8.8 


25.0 


Services and trade 


3.7 


24.1 


15.3 


19.9 


41.2 


Final demand 


22.7 


0.9 


42.7 


67.8 


24.6 



Total 



100.0 



100.0 



100.0 



100.0 



100.0 



Note: Final demand includes, among other items, government and households. 
Note: Final demand output for mining is abnormally low because domestic consumption 
of coal was arbitrarily assigned to real estate and rentals in the service group. 

Source: Bureau of Labor Statistics, Interindustry Flow of Goods and Services by Industry 
of Origins and Destination, 1947 (200-Sector Inter-Industry Table). 



did most of his own selling. 14 These functions are now largely performed by 
other segments of the economy; and the sources of these off-farm inputs must 
have an effect upon agricultural location and thus dilute the influence of the 
immobile resource attributes. 

It is true that agriculture still produces 50% of its non-labor inputs, but 
even here significant functional changes are taking place. Within agriculture, 

"See John H. Davis, "From Agriculture to Agribusiness" Harvard Business Review, Vol. 
34, (Jan.-Feb., 1956), pp. 107-9. 



352 / Regional Distribution of Economic Activities, 1939-1954 

functions are growing less generalized and independent and becoming more 
specialized and dependent. The farmer in the Northeast buys his concentrates 
from the farmer in the Midwest. The Western rancher buys cottonseed from 
a Southern farmer. 

Similar changes in the destination of agricultural outputs have occurred. 
Not more than two generations ago, the typical farm unit consumed most of 
its output. The limited and occasional surpluses were carted to a commodity 
market and usually distributed locally. Currently, something like 26% of 
agricultural output goes directly to services and final demand, 15 and another 
48% directly to manufacturing and processing operations that usually service 
final markets directly. Thus final markets are apt to have a much stronger 
influence on agriculture than on mining, where the outputs that go to manu- 
facturing are usually absorbed into an extensive intermediate series of opera- 
tions. This is made all the more likely when we recall the greater flexibility of 
agriculture on the input side. Further, since intra-industry transfers are im- 
portant in agriculture and not in mining, the influence of extra-industry markets 
is understated for agriculture. 

In general, then, the more purely resource attributes of agriculture are 
probably less dominant in explaining its regional distribution than in the case 
of mining. It is true that agriculture is bound to the land and some sort of 
natural resource endowment is a necessary condition for agricultural production. 
However, it seems plain that this resource endowment is not a sufficient condi- 
tion for agricultural production. Substantial leeway exists for other factors to 
influence the regional distribution of agricultural activity. It must follow that 
market access places a definite stamp on the regional distribution of agricul- 
tural production, not only because it wall make some land or capital substitution 
possibilities more attractive, but also because (through the competition of urban 
uses, etc.) it may make some of these substitution possibilities less attractive. 

Several separate bits of evidence tend to confirm this view. First, as Table 143 
showed, the rank correlation between each state's share of agricultural employ- 
ment and its share of population is .649. (The value of agricultural products 
sold has a slightly higher coefficient — .666) . This compares with a correlation 
of .406 for mining. The distribution of the market (as measured by population 
distribution) therefore apparently has some significance for the location of 
agricultural production — much more so than for mining. This is particularly 
true when we observe that, excluding non-metallic minerals, no sector of mining 
displayed a coefficient much over .200 (see Table 140) . Still the coefficient for 
agriculture is low enough to imply a significant role for resource inputs. 

Second, only 19 states show any form of mining specialization. By contrast, 
only 17 of the 48 states fail to show some form of agricultural specialization. 16 

These two points, however, could be merely a manifestation of the non-local- 

15 Final demand includes the purchases of government and net exports. 

16 I.e., those states whose share of agricultural employment exceeds their share of popula- 
tion, or whose percentage of total employment exceeded the U. S. average. If we include 
those states whose share of farm products sold exceeds their share of population, only 12 
states fail to show some form of specialization. 



Regional Distribution of Agricultural Activity (1) / 353 

ized character of agricultural resources rather than an indication of the impor- 
tance of non-resource factors; so we shall examine two additional bits of 
evidence. 

To make some comparisons, we have used a crude measure of the regional 
distribution of "good" agricultural land. 17 It was determined that 21 states had 
a larger share of good agricultural land than of superior market situations — as 
measured by population distribution. Only four of them, however, showed a 
share of agricultural employment or of value of farm products sold that ex- 
ceeded their share of the good farm land. Most of the concentrations of agri- 
cultural employment (and/or value of farm products sold) relative to share 
of good farm land came in states where the good farm land was in short supply 
relative to the market as measured by population distribution. Further, the 
21 states that had a concentration of good farm land accounted for only slightly 
more than half of the agricultural employment and the value of farm products 
sold. 

This point is also highlighted by a detailed examination, applying multiple 
regression techniques to sample data drawn from State Economic Areas by 
Beverly Duncan, prepared for this study. An effort was made to explain inter- 
regional variations in such items as per-acre value of products sold and rural 
farm population density using three types of explanatory variables — (1) market 
accesses measured by an index of population potential, (2) resource access as 
measured by an index of soil quality, and (3) the effect upon agriculture of 
"rivalry cost" imposed by competition for the land by urban uses, etc., as 
measured by the distance to metropolis and local urbanization. 

The results of this study demonstrate that market access exerts considerable 
influence upon the regional distribution of agricultural activity. The simple 
correlations reveal that the per-acre value of products sold varies inversely 
with distance from a metropolitan center and directly with population potential 
(or closeness to the major centers of population). Inter-area differences in 
this measure of intensity appear to be largely independent of inter-SEA varia- 
tions in the quality of the soil. For the sample of SEA's, 27% of the variance 
in per-acre value of products sold is accounted for by differences in distance 
from a metropolitan center and 16% by differences in population potential. 

Further, such an analysis reveals that the rural-farm density is also directly 
related to population potential. About 29% of the variance in rural farm 
density is accounted for by population potential. Farm population density is, on 
the average, highest in areas of high population potential. Population density in 

"The National Resources Planning Board allocated the total acreage of each state to one 
of five land productivity classes. These were developed by considering ". . . the principal 
physical conditions influencing productivity, such as soil type, topography, rainfall and 
temperature . . ." and assuming ". . . the input of labor and capital will be that most nearly 
capable of maintaining the natural level of productivity, but without irrigation, additional 
drainage, or the addition of lime fertilizer or other amendments except nitrogen fixing 
legumes." For our purpose we have lumped together the "excellent," "good" and "fair" 
grades in contrast to "poor" and "non-tillable." Obviously, there are many limitations to 
our treatment. See National Resources Planning Board, Industrial Location and National 
Resources (Washington: Government Printing Office, 1943) pp. 36-39. 



354 Regional Distribution of Economic Activities, 1939-1954 

the rural-farm sector is, on the average, highest in areas where the physical 
environment is not particularly well suited for agricultural activity. The inverse 
relationship between population potential and soil rating suggests that, on the 
average, the areas with high accessibility to the national market have relatively 
poor natural resources for agricultural production, but nevertheless these areas 
have more persons and more production than would be expected if soil fertility 
was the dominant factor in farming. 

Clearly, modern agriculture is inseparable from the economic functions that 
supply much of its inputs and market much of its outputs. Although a resource 
activity, it enjoys a wide range of substitution possibilities arising out of the 
relative ubiquity of its resource attributes, the alternatives provided by an 
advancing technology, and the external restraints imposed by alternative land 
uses. These substitution possibilities tend to create a fairly wide range of loca- 
tion choices for an industry that makes use of necessary resource inputs. Refer- 
ence to a group of gross indicators has tended to underline this fact. 

We have to recognize, however, that indicators of this type are not fully 
adequate tools for explaining the detailed structure and shifts that characterize 
American agriculture. Changes in market access or input access for specific 
agricultural activities (generated by forces such as changes in demand and 
technology) have stimulated shifts in agricultural activity. The relative flexi- 
bility of agriculture as a resource activity would seem to be reflected in the 
fact that 80 % of the net shift in the value of agricultural products sold is 
generated by differential shifts characteristic of 2-digit agricultural groups. 
While we cannot hope to make an adequate statement of the regional changes 
in access for specific agricultural commodities, it is useful to examine these 
2-digit sectors in some detail. They can make a contribution to our under- 
standing. The next chapter will address itself to this task against the back- 
ground of generalizations touched upon in this chapter. 



22 / Regional distribution of 
agricultural activity: 

(2) THE MAJOR COMPONENTS 



For a deeper understanding of the differential shifts among the states that we 
have observed in agricultural employment, we shall now examine each of the 
major 2-digit components of agricultural activity — field crops, vegetables, 
fruits and nuts, horticultural specialties, dairy products, poultry and products, 
and forest products. We shall also examine at this point the regional influence 
of commercial fishing on the theory that this activity resembles agriculture in 
its general characteristics more than any other resource or processing activity. 



Field Crops 

The field crops sector of agriculture is, without a doubt, the most important 
in explaining the observed changes in agricultural activity. As measured by 
the value of farm products sold, it is the largest of the sectors, accounting for 
40% of the total value in 1954. It is also an important source of inputs for 
the livestock sectors of agriculture and therefore exerts influence on the second 
largest component of agricultural production. 

In addition to its relatively large size, two factors establish the contribution 
of this sector to the proportionality effect upon the regional shift in the value 
of farm products sold in the 1939-1954 period: (1) the rate of growth in 
value for field crops was second only to that for fruits and nuts over this 
period; (2) the areas that specialize in field crop production are almost identi- 
cal with those that show proportionality upward shifts in the total value of 
agricultural products sold between 1939 and 1954. Twenty-seven states showed 
net upward proportionality shifts in total value of farm products sold during 
this period (Figure 60). Of this number, all but four exhibited one or both 
of the following indices of field crop specialization: first, each had a larger 
share of the nation's total field crop production than of the total population; 
second, each had a larger proportion of its total agricultural production in field 

/ 355 



356 / Regional Distribution of Economic Activities, 1939-1954 

crops than the average for the nation as a whole. 1 The four states that failed 
to conform to this pattern — California, Nevada, Wyoming, and Virginia — were 
all except California, states whose shares in the total proportionality effect were 
extremely small. The states with the largest shares of the total net upward pro- 
portionality effect — the Southeast, Southwest, and Plains states — 2 were also 
the states that showed the most marked specialization in field crops. 

While field crops include a widely diversified set of crop activities, some gen- 
eralizations about the locational characteristics of these activities are possible. 

( 1 ) Close direct access to final markets, in relative terms, is not an important 
consideration for field crops as a whole. The 200-sector B.L.S. input-output 
table publishes data for four important field crop sectors: food grains and 
feed, cotton, tobacco, and oil crops. In none of these did the combined output 
going to households, eating and drinking places, and government exceed 3%. 
The tie-in with final markets is almost exclusively through the medium of inter- 
mediate processors. Furthermore, in the case of food grains and feed crops, 
72% of the output is absorbed within agriculture. This can be an important 
factor in strengthening ties with important inputs — particularly the resource 
inputs. 

(2) These resource inputs exert a strong effect in their own right. The same 
input-output table records for these field crop activities a higher proportion of 
total expenditures on land than in the case of any of the other agricultural 
functions reported. These activities are the most land-extensive. Furthermore, 
field crops are the "bread-basket" items in agriculture. They are the basic 
staple commodities that have traditionally made the most stringent demands 
upon the fertility and productive attributes of the soil. Large-scale and effective 
production requires resource inputs of good quality. In this connection it is 
worth noting that the states with the highest concentrations of good farm land 
are in the very regions that show the most concentrated specialization in field 
crop activity. 

(3) Per unit of land area, field crops tend to have a comparatively low- 
value yield. They are less able, therefore, to pay the higher land rents that are 
generated by other uses nearer the final market concentrations. At the same 
time, these products are highly transportable and incur a relatively small 
penalty from distance. 

The combined effect of these factors tends to bring about a concentration of 
field crop activity in areas with poorer market access but favorable resource 
attributes. 

The importance of resource inputs, unencumbered by significant rivalry 
costs, helps to explain which regions can be expected to gain because of the 
contribution made by field crops to the proportionality effect in American agri- 
culture. It also suggests that changes in the access relationship between field 

a The data on which the identification of these states is based are presented in Appendix 
Tables H, and J. 

2 For a discussion of recent changes in the relative importance among crops grown in the 
Plains and Great Lakes states, see John C. Weaver, "Changing Patterns of Cropland Use in 
the Middle West," Economic Geography, Vol. 30, January 1954, pp. 1-17. 



Regional Distribution of Agricultural Activity (2) / 357 




Absolute upward shift 
■jj Relative downward shift 



Figure 62. Net Shift in Value of Field Crops Sold, 1940-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 16.2. Total net shift as % of 1940 value of 
field crops sold: 48.8. Source: Appendix Table J. 



crop activities and their resource inputs may be dominant factors determining 
the major differential shifts. 3 

Figure 62 shows the shift in the value of field crops sold in the 1939-54 
period. The net upward and downward shifts seen here closely parallel the 
total differential shifts shown in Figure 61. Twelve of the fifteen states with 
net upward differential shifts in the total value of farm products show signi- 
ficant upward shifts in field crops. And twenty-nine of the thirty-three states 
with downward shifts in the total have downward shifts in the field crops sector. 

The task of analyzing the contribution of each field crop to the shifts in the 
total goes beyond the scope of our study. "Field crops" is an umbrella term 
that covers a wide range of crop activities. The census lists over sixty cate- 
gories. The most that we can do here with so heterogeneous a classification is 
to point to a few important examples. 

The small grains have, in general, grown in value at a more rapid rate than 
the national average for all crops, and these find their greatest concentration 
west of the 95th meridian. Wheat, oats, and barley all show an above-average 
increase, as does the legume alfalfa; much of this increase has been stimulated 
by military and economic needs during and since World War II. For soybeans, 
the increase has been three times greater than the average; this has undoubtedly 

8 In evaluating these factors, however, due weight must be given to the fact that this is the 
largest segment of agriculture in which price has been supported by governmental action. 



358 / Regional Distribution of Economic Activities, 1939-1954 

influenced the upward shifts shown by Indiana and Ohio. Grain sorghums 
have shown extremely large increases that must have had a considerable effect on 
the upward shift in Texas, New Mexico, and perhaps some of the Plains states. 
And the increase for rice, at twice the national average rate for all crops, must 
have influenced the shift shown by Arkansas. 

On the other hand, some important products have shown below-average rates 
of increase in value. Corn is one of these. This fact must account in large part 
for the declines in the Corn Belt vis-a-vis the Western states and, to an im- 
portant extent, for the downward shift in the Southeast. Peanuts, sweet pota- 
toes, and sugar cane, in all of which the Southeast tends to specialize, have 
also failed to maintain the average rate of increase. So have sugar beets and 
Irish potatoes, which may be a factor in the downward shifts in the mid- 
Mountain states. 

Developments in cotton and in the food and feed grains deserve particular 
attention. 



COTTON 

The over-all field crop shift out of the Southeast and into the Southwest has 
been dominated by cotton. 4 Cotton follows the typical field crop pattern. All 
of its output goes to textiles or foreign trade. Transport costs relative to its 
value have never been sufficient, however, to influence significantly the location 
of cotton culture or of the textile industry that uses its output. It must follow 
that the shift out of the Southeast has taken place because of a change in that 
region's access to important resource inputs for cotton culture relative to the 
Southwest and California. The story of this change is one of the most dramatic 
in American economic history. 

The Southeast was settled to open up new cotton and tobacco lands because 
these crops were highly profitable and because the foreign credits earned 
through their export were important to a young debtor nation. The conse- 
quence of this was a high degree of crop specialization at an early period that 
did not become characteristic of other agricultural functions until quite modern 
times. Early in history the Southeastern farmer had to develop intimate input- 
output relationships with off -farm functions — again a relationship that de- 
veloped in other farming systems and regions only recently. This system of 
specialization and off-farm ties created four significant institutional patterns. 
First, cotton was very labor-intensive, requiring enormous amounts of hand 
cultivation. This was a serious restraint against its production in a young 
labor-hungry country. The consequence was slavery. Second, with the break-up 
of the plantation system, the labor-intensive character of cotton culture forced 
a fragmentation of land holdings into small units. Third, because cotton could 
not be pilfered for food and was worth little until ginned, its cultivation lent 

4 The farming system is heavily oriented to the resource characteristics of the land and 
climate. Cotton production requires 200 frost free days (which limits it to the areas south of 
the 36th parallel) and makes rather specific requirements upon soil, topography and 
moisture. 



Regional Distribution of Agricultural Activity (2) / 359 

itself to absentee ownership and created a region addicted to share-cropping. 
Fourth, this fact, combined with the importance of off-farm input and market 
sources, created a dominant role in Southeastern agriculture for the merchant- 
banker as well as the absentee owner. 

Add to these cultural and institutional relationships the fact that resource 
exhaustion became important. Cotton in certain circumstances not only ex- 
hausts soil nutrients but vitiates humus. In the Southeast, at best, this resulted 
in enormous increases in fertilizer applications and, therefore, farm costs. At 
worst, it created wholesale erosion and farm abandonment. 5 

Most of these factors came to have a special significance with the advent of 
mechanization. 6 Because mechanization requires even topography and rock-free 
soil, large land holdings, and capital in large amounts, its introduction altered 
the character of the inputs essential to cotton culture. 

In short, a number of changes reduced the Southeast's comparative advantage 
in cotton production and augmented that of the West. (1) Significant soil 
exhaustion had taken place in many parts of the Southeast. (2) Uneven topo- 
graphy and small land-holdings denied the gains of mechanization to many 
Southeastern cotton-producing areas. (3) The history of cotton culture in the 
Southeast explains why the Southeast had relatively limited access to the neces- 
sary capital inputs. (4) The spread of irrigation increased the West's access 
to some extraordinarily fertile cotton lands. 7 

The significance of cotton to regional economic growth, however, merits 
consideration in terms beyond those involved when the focus is on changing 
access and consequent shifts in field crops. The economic conditions which 
created the cotton culture of the Southeast, and the institutional arrangements 
that evolved to facilitate it, have created a regional maladjustment so great 
that it affects the growth prospects for the whole Southeastern portion of the 
nation. 8 The phenomenon of erosion and the westward shift of cotton culture 
are making imperative a new form of agricultural production, particularly a 
new form of grass, grain, and livestock culture. This, however, calls for in- 
creased mechanization and larger land holdings. Unfortunately, the institutional 

6 The crop was intertilled year after year on hilly uplands that were particularly susceptible 
to erosion. In addition, in most instances no winter cover crop protected the land after the 
crop was removed. Little or no organic matter was returned to the soil as plant residues or 
animal manure. 

6 If cotton was one of the first farming systems to become specialized, it was one of the last 
to become mechanized. The peculiar nature of the cotton plant and the specialized labor 
requirements it makes during the growing season created mechanical and technical problems 
not found in wheat or corn. Currently most of these problems have been solved and a wave 
of mechanization has set in. 

7 Another factor has been the boll weevil, which reached the height of its destructiveness in 
1921, ruining at least one-third of the cotton harvest in that year. The famed sea-island 
cotton industry along the Atlantic Coast never fully recovered from damage inflicted by this 
insect, which is particularly effective in moist climates. In contrast, the drier climates of 
northwestern Texas, western Oklahoma, and southern California tend to repel the insect, 

8 Cotton has been significantly aided in this bit of mischief by tobacco which has most of 
the same characteristics outlined above. It will not serve our purpose, however, to expand 
this discussion further. 



360 / Regional Distribution of Economic Activities, 1939-1954 

arrangements evolving from a century of cash cropping inhibit the needed 
transition. The southern farmer owns little capital, while the southern merchant- 
banker, habituated to the security of a staple commodity, hesitates to risk new 
credit and tenant arrangements. As a result, many of the southern farmers 
have not been able to finance the drastic changes which effective adjustment 
of the Southern economy would require. The instability of cotton prices over 
generations and the cost-price squeeze growing out of the aforementioned fac- 
tors have been hard on the man on the land. The resultant human erosion has 
been more serious than that of the land. 

Although rational adjustment has been slow in coming to the Southeast, it has 
been taking place and the mere fact of adjustment, even if limited, is creating 
new problems related to the Southeast's growth. Cotton and tobacco are very 
labor-intensive operations. The new, more rational farm systems require much 
larger holdings and much less labor. This creates in effect an area with a large 
surplus labor supply and significant underemployment. A population adjusted 
to an outdated cotton culture is now too large for its land resource base. This 
creates a situation where people must either move out or alternative employment 
move in. Both these movements have been taking place, but not fast enough to 
offset regional inequities in welfare. 



FOOD AND FEED GRAINS 

Food and feed grains share the field crop characteristic that little output goes 
to final demand and land costs represent a relatively large part of total inputs. 
They share therefore the general feature of being oriented to the resource 
characteristic of the land. In fact, this orientation is probably strengthened by 
the fact that three-fourths of the output of these grains is absorbed by agricul- 
ture itself — principally by the livestock sectors. 

Corn and its livestock ties explain an important part of the shifts shown in 
Figure 62. Corn is grown successfully in almost every part of the United States. 
It really flourishes, however, in the region between the 98th meridian and the 
Appalachians above the Ohio River. The long hot summer days and hot nights, 
combined with top grade soils, have created an historic dominance here for this 
crop. Mechanization, new fertilization techniques, hybrid corn, and other tech- 
nological advances have only seemed to accentuate the relative land-resource 
advantage of this area. 

Why, then, has this area displayed a net downward shift in field crops? Iowa 
shows the most marked downward shift of any state and certainly it is the corn 
state. Here the data — i.e., using the value of farm products sold as a measure — 
turn out to be somewhat misleading. Actually in physical terms, these Mid- 
western states would be included in the upward shift category, as compared to 
the rest of the country. The answer lies in the fact that most corn is not sold 
but is converted into prime beef and pork. During the 1939-54 period a higher 
proportion was used directly for feed rather than sold. This shows up in Figure 
62 as a net downward shift. 

This factor also helps to explain the upward shifts exhibited by Indiana and 



Regional Distribution of Agricultural Activity (2) / 361 

Ohio. Here just the reverse has happened. Less feed grain was fed and more 
sold. The reason for this is the significant industrial expansion that has taken 
place in these states. Large numbers of farms in these areas have become part- 
time, with the operator working off the farm. Under these circumstances he 
can raise the grain for sale but he cannot administer a feeding program. 

Wheat and grain sorghums played an important part in the West's relative 
gain in field crops during the period under review. 9 Two factors have been 
favorable to its position in wheat: 

(1) Changes in technology, such as techniques of dryland farming and the 
development of drought-resistant strains of wheat, as well as the development of 
large machinery, seem to have favored the West; and (2) factors favoring the 
expansion of corn culture in the Midwest have resulted in some displacement of 
wheat in the transition area between the two cultures. (This is a good example of 
rivalry costs transmitting the influence of change in one crop culture to 
another.) Add to this development in wheat the advent of grain sorghums, 
which have shown a marked expansion in the West, and much of the shift 
configuration observed is explained. 

The upward shift shown by Indiana and Ohio is attributable in part to soy- 
beans, which have tended in these states to replace crops with a smaller output 
per acre. 10 Soybeans have also been responsible for the relative decline of 
competing oil crops, such as peanuts. 



Livestock Products 

Livestock products constitute the second most important sector of the agricul- 
tural economy. In 1954 they accounted for almost 29% of the total value of 
farm products sold. Like field crops, they are in the rapid-growth category. It 
is not surprising, then, that the states specializing in their production — the 
Mountain and Plains states and some of the Great Lakes states — show an upward 
proportionality shift in the value of farm products sold (see Figure 60) . 

Three generalizations can be made about the locational factors that determine 
the distribution of livestock production: (1) The largest single input tie for 
the livestock products sector is with other agricultural sectors; over 50% of its 
inputs come from agriculture — particularly the feed grains and forage sectors. 
Access to high-quality land resources is not directly important as in field crops. 
But access to the feed sectors is an important locational determinant unless offset 
by the influence of some other factor, such as market access. (2) Less than 
10% of the meat output goes to final demand; 81% goes to intermediate proces- 
sors, chiefly packing houses. Since this intermediate processing involves con- 

9 The proportionality shifts recorded came about because of the change in relative values 
stemming from the war-inspired increase in demand and the politically inspired postwar 
price support program. Wheat output declined steadily from 1919 to 1939. With the war 
came good prices and a concerted effort to increase production. Wheat benefited from the 
increase in demand more than most agricultural products. 

10 This has been less true in Iowa and the western end of the corn-soy belt because there 
it has tended to replace corn as a result of the federal acreage-control program. 



362 / Regional Distribution of Economic Activities, 1939-1954 

siderable weight loss, the tendency is for the packing houses to locate near the 
livestock sources, thus increasing the significance of access to basic feed inputs. 
(3) There is another factor that might be expected to offset this tendency to a 
degree. Livestock products are high-value products and are relatively expensive 
to transport either dressed or on the hoof. Ordinarily this would incline an 
industry to a location toward the market. However, this factor has not been able 
to overcome the importance of feed inputs and the weight-loss characteristic of 
intermediate processing. But it has been sufficiently important to bring about 
some degree of regional specialization in feeder-stock and finish-feeding, with 
finish-feeding done as close as possible to the source of high-grade finishing 
feeds and to the packers. 

Actually, the whole activity of livestock products production looks somewhat 
more like a manufacturing activity than a resource activity. The animal, in 
effect, is a biological machine that takes purchased and farm-raised inputs and 
produces a marketable product. As a matter of fact, often (as in dry-lot feed- 
ing) the significance of land as an input rests more upon its situs than its 
resource characteristics. The most economical position for the animal depends 
upon the relative costs of assembling inputs and marketing outputs. In this 
case, the input sources dominate the location scheme. It is interesting to note 
in this connection that the correlation between livestock products and population 
is very slight (Table 144). 

Figure 63 shows the net shifts in the value of livestock products sold during 




upward shift 
Tj Relative downward shift 



Figure 63. Net Shift in Value of Livestock Products Sold, 1940-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 15.1. Total net shift as % of 1940 value of livestock 
products sold: 41.5. Source: Appendix Table J. 



Regional Distribution of Agricultural Activity (2) / 363 

the 1939-54 period. The major upward shifts over these years were in the Corn 
Belt states, California, and the Southeast. This configuration suggests that live- 
stock products are quite an important factor in the total differenial shifts shown 
in Figure 61. Of the fifteen states that show net upward differential shifts in the 
value of farm products, twelve also show net upward shifts in the value of 
livestock products sold. It is important to note, however, that the center of the 
Corn Belt and certain of the Southeastern states show downward shifts in agri- 
cultural products as a whole despite upward shifts in livestock products. 

The principal locational factors that have prompted the differential shifts in 
the production of livestock products appear to be associated with regional 
changes in access to feed. This is certainly the dominant factor behind the 
largest net upward shifts — those in the corn-soy belt. Hybrid corn has been 
influential in expanding the productivity of this region. So also has been the 
increasing practice of finish-feeding, which places a premium on access to finish- 
feeds such as corn. 

The Southeast's upward shift is also associated with a relative increase in its 
access to feed sources. This is another part of the story of adjustment in this 
region away from the row-cropping systems that have debilitated the soil. 
Rational adjustments in the Southeast call for an extension of cover-cropping, 
and this expansion of potential livestock feed encourages an expansion of 
livestock activities. 11 

Most of the Southwestern and Mountain states display net downward shifts. 12 
These states are livestock specialists. They appear, however, to have suffered 
relative losses in input access. The factors here are as follows: (1) These states 
are predominantly feeder-stock areas and the increased practice of finish-feeding 
means that they lose out relatively in a measure based upon agricultural values. 
(2) Much of the mid-Mountain region has been overgrazed, and this practice 
is leading to a deterioration in access to feeds. (3) Competitive land uses may 
have occasioned some loss of access to feeds in Texas and New Mexico. Cotton 
and grain sorghums, for example, have made striking gains in acreage in these 
states. 

The significant upward shift in California is of particular interest because the 
principal reasons for this gain seem to have been a large increase in the regional 
market. The remoteness of the Far West region tends to place it outside the 
dominant national transfer patterns for livestock products. As a result, the 
regional market exerts a stronger modifying influence than is usual for the 
nation as a whole. 



"Technology has aided this shift also by introducing new pesticides and strains of stock 
(the Brahmin) that are better adapted to warmer climates. The adjustment is going slowly 
because it requires heavy capital investments. Also the intermediate processing units are 
not as readily available in the Southeast, and the institutional factors mentioned earlier act 
as restraints. However, there seems to be a definite adjustment under way. 

12 In the case of Texas, it is particularly striking and serves to explain the discrepancy 
between the strong upward shifts in field crop activity in the face of a downward shift in 
the total value of farm products. It should be noted that Texas cattle sales were unusually 
low in 1954, due to a drought. The trends discussed above are basically secular, however. 



364 / Regional Distribution of Economic Activities, 1939-1954 



Dairy Products 

Dairy products in 1954 accounted for 13.5% of the total value of farm 
products sold. The rate of increase for this sector between 1939 and 1954 was 
smaller than for any other sector of agricultural activity. As a result, the states 
specializing in dairy products correspond closely with those shown in Figure 
60 as having experienced net downward proportionality shifts over these years. 
Twenty-eight states show a specialization in dairy products as measured by one 
or both of the following indices: (1) a larger share of the nation's dairy prod- 
ucts production than of its population; (2) a larger proportion of their total 
agricultural production in dairy products than the average for the nation as a 
whole. 13 Of these twenty-eight states, only nine fail to show downward propor- 
tionality shifts in the total. The areas of specialization are in the New England 
and Middle Atlantic regions, the northernmost Lakes and Plains states, and the 
Far West states plus Utah and Idaho. 

In the production of dairy products, as in livestock production, the animal is 
essentially a biological factory, and the most economical location for it depends 
upon the relative costs of assembling inputs and marketing outputs. With 43% 
of the inputs coming from feed sources and 58% of the outputs going to final 
markets, one can see that the dairy culture might be torn between its affinity for 
feed sources and its affinity for markets. This is precisely the dominant loca- 
tional fact about this production system. 14 

The distribution of dairy products specialization reveals a split in production 
orientation along these lines. The area of concentration in New England, the 
Middle Atlantic states, and the Far West is essentially explained in terms of 
markets. The big concentration in the Great Lakes States, however, is hard to 
rationalize on the basis of market proximity. The production statistics show 
that the densely populated New England and Middle Atlantic areas market fluid 
milk, with almost 100% going to final demand, while the Great Lakes region 
markets fluid milk and cream primarily to processors to be transformed into 
cheese, butter, and condensed and dehydrated products. 

The reasons for this split are fairly logical. Fluid milk for comsumption is a 
highly perishable product that loses no weight in processing and requires special 
handling. Food concentrates can be shipped at low rates to supplement local 
feeds so that they can be grown on relatively inferior lands. In contrast, the 
production activity in the Great Lakes region is oriented to the resource charac- 
teristics of the land. The area is too far north for the best corn culture and too 
hilly to make good crop land. On the other hand, grass and forage crops thrive 
almost better than any place else in the United States. It is a favorable low- 
cost area for inputs. If the outputs are transformed into less perishable and 
more transportable products, this favorable resource advantage can be exploited. 

13 The data on which the identification of these states is based are presented in Appendix 
Tables H and J. 

"Production and distribution are influenced by trade barriers in the form of "health" 
restrictions, but it is difficult to evaluate this factor in the present context. 



Regional Distribution of Agricultural Activity (2) / 365 

Figure 64 shows the shifts that occurred between 1939 and 1954 as measured 
by value. The Southeast and three other states — Wisconsin, Pennsylvania, and 
California — made the greatest relative gains. 




Absolute upword shift 
{/■■v'l Relative downward shift 



Figure 64. Net Shift in Value of Dairy Products Sold, 1940-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 11.8. Total net shift as % of 1940 value of 
dairy products sold: 23.3. Source: Appendix Table J. 



The dairy products shifts tend to run counter to the total differential shifts. 
Out of twenty-two states showing upward shifts in dairy products, only eight 
also show upward shifts in total value. This can be partly accounted for by the 
proportionality effect just discussed. The gain shown in the Middle Atlantic 
states is a consequence of the use of the value measure; it seems to result pri- 
marily from the fact that the fluid-milk areas were less affected than the milk- 
processing areas by the drop in the price of butter. The decline in the Great 
Lakes states, with the exception of Wisconsin, seems to be tied in part to the 
declining price of butter and in part to the rivalry costs imposed by an expand- 
ing corn-soy culture. The fact that the technology of dairy raising increases the 
economy of larger units tends most to favor Wisconsin, where the ecological 
factors are the most favorable. The decline in the Plains states may be tied to 
the price of wheat, which gives the farmers in that area a more favorable 
alternative economic opportunity. 

The growth in dairying in California seems to be most closely tied to the 
growth of the market. Milk production here is oriented more to direct marketing 
than to processing, and the share in the nation's dairying activity has increased 
less than the share in total population. 



366 / Regional Distribution of Economic Activities, 1939-1954 

The shift in production into the Southeast seems to be related to considera- 
tions of both market and resource access. The Southeast has always been a 
deficit supply area for milk. Further, per capita consumption in the Southeast 
has always been low. With rising incomes, there is an expanded use of milk 
products here. On the supply side, the expansion of dairying is closely tied to 
the decline of cotton and other crops in the Southeast. This expansion has been 
called for in part because it represents a favorable income alternative to a 
declining system. It has also been called for because the grasslands that are 
associated with dairying are the only sensible alternative to abandonment of 
many of the more eroded and overworked lands. 



Poultry Products 

Poultry products, which accounted for about 8% of the total value of farm 
products sold in 1954, show a proportionality effect closely resembling that of 
dairy products. 

The conflict between the pull of the market and the pull of feed sources is 
seen here also; 78% of the output goes to final demand and 76% of the inputs 
come from agriculture. Measuring specialization for this sector in the same 
way as in the earlier sectors examined, we find three separate areas of speciali- 
zation — (1) the Northeast (including New England and the Middle Atlantic 
states), (2) the Far West and Utah, and (3) a few of the Southeastern and 
Plains states. The first two areas appear to be closely associated with market 
concentrations and the last appears to be more closely associated with good feed 
access. The first two, it is also interesting to note, tend to specialize in fresh 
poultry meat and the last in eggs and canned poultry. 

Figure 65 shows the shifts that occurred between 1939 and 1954 in the value 
of poultry products sold. The major upward shifts have been in the Southeast, 
the Atlantic Coast states, and California. 

Basic to all of these shifts is the development over the last 15 years or so of 
the broiler industry as a separate producing activity. Previously, poultry raising 
was a by-product of joint-cost operation carried on with other agricultural 
activities. The technology of poultry raising has advanced so strikingly that now 
the most economical method of production is a fairly large-scale specialized 
operation. 15 This technological revolution in the poultry industry has had two 
consequences. (1) The industry now relies more heavily on prepared feeds and 
is better adapted to large-scale dressing and processing. This has tended to 
strengthen the relative importance of market access. (In this connection, note 
the gains in California and the Northeast.) (2) Under the new system, poultry 
raising becomes more labor-intensive. Access to adequate cheap labor appears 
to have assumed some importance. The expansion in Maine seems to be associ- 
ated in part with the market but mainly with the need for an alternative outlet 

w The technological advance has proceeded side-by-side with vertical integration of the 
industry, which has resulted in increased capital availability and risk reductions to the 
farmer. 



Regional Distribution of Agricultural Activity (2) / 367 




:';■' I Relotive downward ihrft 



Figure 65. Net Shift in Value of Poultry Products Sold, 1940-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 23.8. Total net shift as % of 1940 value of 
poultry products sold: 53.4. Source: Appendix Table J. 



for part-time fishing employment. This has been particularly feasible because 
fish and poultry are usually marketed through the same channels. The great 
expansion in the Southeast is also partly motivated by the needs for labor 
outlets and partly by the increasing levels of living. 



Fruits and Nuts, Vegetables, and Horticultural Specialties 

These three sectors of agricultural activity combined account for about 11% 
of the total value of farm products. The first of them — fruits and nuts — is a 
rapid growth sector; the other two are slow-growth sectors. All together, 
however, they comprise such a small part of the total that their proportionality 
effect is not likely to be very significant. Measuring specialization by the same 
techniques as before, 16 we see that these agricultural activities all show the same 
areas of concentration — (1) the highly populated Northeast (particularly the 
Atlantic coast and the Great Lakes states), (2) Florida, and (3) the Pacific 
coast states. 

These states also display similar input-output structures that contrast with 
those displayed by field crops and animal products. Nearly three-fourths of the 

16 Utilizing the data in Appendix Tables H and J, we identify all of the states that have 
a larger share of these products than population and those states for which these products 
as a percentage of total agricultural production for the state exceeded the national average. 



368 / Regional Distribution of Economic Activities, 1939-1954 

output goes to final demand categories and only one-fourth to processing opera- 
tions. Practically none is traded within agriculture itself. Labor is the major 
input item. The expenditures for other items are so widely distributed that these 
inputs probably have no significance as locational factors. The products are 
in general high-yield and high-value products that bear heavy transfer costs. 

As might be expected, then, these activities are highly associated with centers 
of market activity. This is especially true for horticultural specialties. 17 Of all 
agricultural activities this sector reveals the highest rank correlation with popu- 
lation (see Table 143). Not only are these activities seen to have been highly 
associated with population in the 1939-54 period, but the shift pattern they 
exhibit is highly associated with the shifts in population over these years. 18 

The fruits and nuts sector and the vegetables sector, however, are considerably 
further down the scale of rank correlations with population. Figures 66 and 
67 show the shift picture for these two sectors. Although this pattern does not 
correspond closely with the total differential shift (Figure 61), these sectors 
were unquestionably influential in generating the total differential shifts in 
Washington, California, Arizona, and Florida. 



FRUITS AND NUTS 

Some 88% of the net upward shift in fruits and nuts production was claimed 
by the three states — Florida, California, and Washington; in the main market 
areas, fruit production barely held its own in terms of the national average. A 
number of factors would seem to be operative here: (1) Citrus is the rapid- 
growth component of this sector, and none of the main market areas can meet 
the major resource requirement of citrus production — a frost-free climate. 
(2) Fruit production in general has always favored moderate climates. The 
problem of perishability, however, in the past gave overwhelming advantage to 
good market access. But improvements in the technology of storage and trans- 
port have offset the market advantage and permitted production to seek a greater 
resource advantage. Frozen foods and concentrates offer methods of food pres- 
ervation with fresh fruit quality. Air transport has been successfully used in 
marketing prime-grade fruits and vegetables, though not yet in significant 
volume. (3) Florida has for some time been showing gains relative to Cali- 
fornia. Its production is cheaper and it is closer to major markets. In recent 
years also California output has been hurt by the fact that many of the good 
citrus-producing areas in Los Angeles county were being forced out of produc- 
tion by the competition of urban and suburban uses for the land. 

VEGETABLES 

Except for Delaware, which showed a small relative gain in vegetable produc- 
tion, all of the areas closely allied with the major markets suffered significant 

"Horticultural specialties consist of such things as nursery products, cut flowers, vege- 
tables grown under glass, etc. 
18 See Appendix Tables H and J. 



Regional Distribution of Agricultural Activity (2) / 369 




Absolute upward shift 
| ?H Relative downward shift 
| ^J Absolute downward shift 



Figure 66. Net Shift in Value of Fruits and Nuts Sold, 1940-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 16.2. Total net shift as % of 1940 value of 
fruits and nuts sold: 49.5. Source: Appendix Table J. 




Absolute upword shift 
''. | Relative downward shift 
I Absolute downward shift 



Figure 67. Net Shift in Value of Vegetables Sold, 1940-54. 

State figures represent % of total net shift. Total net shift as % of 
incremental change : 27.5. Total net shift as % of 1940 value of 
vegetables sold: 61.5. Source: Appendix Table J. 



370 / Regional Distribution of Economic Activities, 1939-1954 

relative losses between 1939 and 1954. Florida, California, and Arizona in 
particular and the West in general had the major gains. Wisconsin and Minne- 
sota also showed upward shifts, based on an increased production of vegetables 
for canning. The factors responsible for the shifts in vegetable production are, 
in the main, similar to those accounting for the shifts in production of fruits and 
nuts. One factor that has been of special importance, however, is irrigation. 
Effective moisture control has been a great relative advantage to the states that 
show the greatest upward shifts — Florida, California, and Arizona, Storage and 
transport innovations have made the exploitation of these advantages feasible. 

One comment of general import deserves comment in this connection. Greater 
latitude is found in the use of immobile agents when existing techniques allow 
for mobile substitutes in the form of capital. We find an example of this in the 
case of horticultural specialties. They tend to be grown close to markets irre- 
spective of land resource attributes. This is made possible by substituting capital 
(in the form of greenhouses, etc.) for land. It is important to realize, however, 
that the substitution of mobile capital for land may increase the immobility. In 
some cases the purpose of the capital expenditure is to release an important 
latent land resource rather than minimize the importance of land in the produc- 
tion complex. Such is the case with irrigation. Those lands that were irrigated 
first are those that could release the maximum potential land productivity. The 
consequence is that vegetable culture, which benefits particularly from irrigation, 
is becoming less population-oriented, not more so (see Table 143). 



Regional Influence of Forest Activity 

One of the 2-digit components of the value of farm products is farm forest 
products. We have chosen, however, not to discuss the influence of forest activ- 
ity in that context for two reasons. (1) This is one farm activity where an 
important part of the output for the sector as a whole comes from off-farm 
sources. To concentrate upon farm forests would present a distorted picture of 
the total. (2) This is a sector for which we have quite good data on employ- 
ment and it is not necessary to limit ourselves to value data. We discuss here 
the influence of total forest activity — both farm and off-farm. Logging employ- 
ment is our major focus. 

In terms of employment, logging and forest activities account for only about 
a third of 1% of total employment in the nation. Its growth, as measured by 
employment increases, is about par with the average national increase in employ- 
ment; the proportionality effect, therefore, is not significant. As with certain 
other activities, however, forest products become especially important for under- 
standing the performance of particular regions. 

Forest activity, like agriculture, is ubiquitous in the sense that there is some 
activity in every state. Only three areas, however, have sufficient production 
potential to be classified as surplus supply areas — the Far West, the Southeast, 
and Upper New England. 19 

19 In an earlier period, the Great Lakes states would have belonged in this array. But de- 
clines in production and increases in population bring them now to the point where their 



Regional Distribution of Agricultural Activity (2) / 371 

Logging activity, like mining, must take place where substantial reserves are 
found. Market access does not play a very important role regionally in its 
location. In the first place, three-fourths of the total output of logging goes to the 
mills that produce pulp, lumber, veneer, and plywood. Because of the weight 
loss associated with processing logs, these manufacturing activities are located 
close to forest reserves. 




Absolute upward shift 
v">] Relative downward shift 

| Absolute downward shift 
* Even 

Figure 68. Net Shift in Logging Employment, 1940-50. 

State figures represent % of total net shift. Total net shift as % of 
incremental change: 51.4. Total net shift as % of 1940 logging 
employment: 17.8. Source: U. S. Bureau of the Census, U. S. Census of 
Population 1940, Vol. Ill, labor force data in State Table 11; and U. S. Census 
of Population 1950, Vol. II, labor force data in State Table 73. 



share of production is less than their share of the population. There is an interesting kind 
of specialization among these regions. The Southeast tends to specialize in pulpwood and the 
Far West in sawlogs for lumber. This is due mainly to differences in the sizes of standing 
timber. The old-growth timber in the Southeast which has been cut over decades ago for 
sawlogs has been replaced by rapidly growing but lower quality second-growth pines and 
hardwoods of smaller sizes, primarily valuable for pulp-making. In the West, the remaining 
extensive areas of large old-growth timber can be converted more efficiently and profitably 
into high-value lumber. Nevertheless, in terms of the volume produced, Southern pine lumber 
is next in importance to Douglas fir of the West. Another factor, institutional in nature, will 
probably strongly influence the future trends in sizes and disposition of timber produced by 
both regions. In the Southeast, and to a lesser degree in New England and the Great Lakes 
states, timber holdings are quite small on the average and vested in individual proprietors. 
In the Far West large national forests and corporation holdings are more common. It is a rare 
individual owner with small holdings that has the knowledge and capital resources to place 
into effect the long rotations necessary for sawlogs. 



372 / Regional Distribution of Economic Activities, 1939-1954 

This does not prove, of course, that the market could not influence the whole 
complex of logging and processing activities. Historically such an influence has 
been important because historically forest reserves were large relative to current 
requirements and logging was conducted as an essentially mining type operation. 
Virgin forest reserves were exploited first in New England, and forest activities 
then moved westward to the Great Lakes states, southward to the Southern states, 
and finally to the Far West. This sequence of movement was influenced to a 
considerable degree by access to the major eastern markets. In our current 
period, however, these market considerations do not exert much influence upon 
the location of logging activities. There are no longer any new bodies of timber 
that are not being exploited because of distance from the market. All of the 
principal submarginal (costwise) forest lands of yesterday have now passed into 
use and timber resources are in short supply everywhere. Currently timber is 
exploited where it can be found. 20 

It follows, therefore, that most regional shifts in forest activity must result 
primarily from changes in access to forest resources. Figure 68 records the 
shifts in logging employment from 1940 to 1950. 21 Several major changes in 
the pattern of activity are apparent. (1) In the Far West, Washington has 
experienced a marked absolute decline and accounts for almost 38% of the 
total net downward shift. A shift out of Washington and into Oregon and 
California is evident. (2) There is a marked upward shift in the Southeast. 
These states combined account for 85% of the total net upward shift (Florida 
is a striking exception.) (3) New England and the Great Lakes states show 
slight absolute and relative downward shifts in activity. 

The changes in access to forest resources to which these shifts bear witness 
stem from a number of factors. Perhaps the most basic of these is resource 
exhaustion. This is central to the current as well as the historic declines in New 
England and the Great Lakes states. It is almost the total explanation of the big 
declines in Florida and Washington; the forest reserves here have been cut out 
to the point where they will no longer support the previous level of production. 
These declines forced expansion in the only areas where under-utilization of 
reserves existed. Northern California and Oregon have the last remaining stands 
of virgin timber. At the same time, second-growth timber has been coming back 
in volume in the Southeast. 

Closely associated with this phenomenon of exhaustion has been a change of 
emphasis in forest activities. There has been a perceptible shift away from a 

20 Of the non-resource inputs, labor accounts for about 75% of the total and the balance 
is spread between manufacturing and services without any concentration upon a specific 
input. There is a special provision in the minimum wage act that exempts logging crews 
of twelve men or less. This has led, in the Southeast, to the widespread use of contract 
loggers utilizing small crews which are recruited in a surplus labor market at very low 
wage rates. This has tended to give the Southeast an advantage in cost. It does not appear, 
however, that this has been a major factor in determining the regional distribution of 
activity. 

21 Data for logging activities that coincide with our period (1939-1954) are not available. 
This information is as close as we can come and is based upon occupation data in the 
census of population. 



Regional Distribution of Agricultural Activity (2) / 373 

mining type of activity toward something approaching silviculture and a hus- 
banding of the forest reserves. This means that timber cut will be brought 
more and more into line with realizable growth. 22 

A number of additional factors have contributed to the over-all picture. A 
change in the composition of demand for forest products has played a role. 
From 1939 to 1954 the demand for pulp and paper products increased by about 
twice as much as the demand for lumber and lumber products. This factor 
favored expansion in the Southeast, which is heavily oriented to the production 
of pulp logs. Technological change has aided this shift. The development of the 
sulphate process for producing paper pulp brought the resinous Southern pines 
into use for this purpose. 

Of the growth states, only California and Oregon gained any support from 
forest activities during the 1939-54 period. In the major decline states, the 
falling-off in forest activity intensified the total downward tendency. The excep- 
tions are the states in the Southeast. Here expanding forest activities seem to 
be aiding the necessary adjustments in the agricultural sector of the economy. 
However, they do not provide a very high income alternative, nor is the employ- 
ment alternative sufficient to go very far in absorbing the agricultural surplus. 23 



Regional Influence of Commercial Fishing 

Commercial fishing plays a relatively small part in regional growth. By the 
most liberal employment estimates, that include casual as well as full-time 
workers, there are less than 150,000 wage jobs involved. The shifts that have 
occurred in this industry, however, are almost as complex as those in any other 
field of activity. 

The coastal states claim 90% of the total fishing employment; most of the 

"The Timber Resource Review has this to say: "A comparison of realizable growth and 
1952 timber cut by sections . . . shows that the current distribution of the timber industries 
is not related to timber growing capacity. The West, which is supplying 46 per cent of the 
timber cut, has only 29 per cent of the realizable growth. On the other hand, the Southeast, 
which is producing 40 per cent of the timber cut has 45 per cent of the realizable growth. 
The North contributes only 14 per cent of the timber cut, but has 25 per cent of the real- 
izable growth ... in the long run, output by regions will tend to become adjusted more 
closely to realizable growth than is the case today. This will mean a reversal of recent trends 
and a shift of a larger proportion of the timber cut from West to East." See C. E. Behre and 
S. B. Hutchison, "Timber Supply and Quality of Domestic Timber," Timber Resource 
Review (Washington: U. S. Department of Agriculture, Forest Service, 1955), Chap. 7, p. 6. 

^Something can be said about future prospects for this industry. In the Far West, 
California and Oregon may soon join Washington in decline. The timber cut in these 
areas is considerably beyond realizable growth as the last virgin stands in the country are 
being exploited. This area will suffer both an absolute and relative decline until it arrives 
at an adjustment with its long-run sustainable yields. The activity in the Southeast will 
probably make a small additional gain in absolute terms and a much larger one in relative 
terms. The Great Lakes and upper New England states appear to be in for a reversal of 
trend. Low current timber cut relative to realizable growth, and a new semichemical process 
that allows low grade hardwoods to be used for pulp seem to indicate a revival of forest 
activity in these areas. 



374 / Regional Distribution of Economic Activities, 1939—1954 

remaining 10% is centered in the Great Lakes states. These coastal states, 
however, do not share in fishing production and employment in proportion to 
their access to open water. The sea, no less than the land, has its fertile and its 
barren regions. Commercial fisheries are all narrowly localized and depend 
upon a favorable combination of oceanographic factors. As a consequence, four 
major fishing regions can be identified — coastal New England, the Chesapeake 
area, the Gulf states, and the Pacific Coast states. 

Figure 69 shows the principal shifts that have taken place within the fishing 
industry during the 1940-50 period. Maine, California, and the Gulf states 
(particularly Texas and Florida) have experienced the major expansions over 
these years, and Virginia, Maryland, Michigan, and Mississippi, the major 
declines. 24 

Changes in access to markets and to non-resource inputs do not appear to 
have been a significant factor in these shifts. Over 50% of the total catch finds 
a market in the form of meal and oil, and that market is not influenced by the 
regional shifts of population that have taken place. Most of the balance is 
canned or frozen at the port of entry. The marketing of fresh fish, where market 
access really becomes important, is a very small and declining segment of the 
total activity. On the input side, the labor inputs and capital in the form of 
boats and nets do not appear to exert any regional influence. It follows that the 
shifts result primarily from changes in access to the fishing resource. 

A number of factors have played a role in changing regional access to fishing 
resources. Resource exhaustion is important here as in other resource activities. 
It is particularly important in explaining the decline in Washington and Oregon. 
The salmon fisheries do not yield as large a catch as formerly. 25 Resource 
exhaustion has also been an important factor in explaining the declines in the 
Great Lakes and Mississippi fisheries. In the Great Lakes region the most direct 
cause has been an increasing infestation with lamprey which prey on the com- 
mercial fish. In the Mississippi area resource exhaustion results from stream 
pollution and overfishing. The decline in Massachusetts is also directly tied to 
a decline in the groundfish catch that has been the mainstay of its fleet. 

At the same time, Maine, the Gulf states, and California have had an increase 
in their access to fishing resources. A combination of circumstances is respon- 



24 The shift data indicate a major decline for the Chesapeake area. This decline appears 
to be spurious, however. Similar data supplied by the Fish and Wildlife Service of the 
Department of the Interior show this area to be holding its own in employment and a 
further check against production figures fails to substantiate this downward shift. The 
Fish and Wildlife figures appear to be consistent with the rest of the picture. The difference 
here appears to be due to statistical discrepancy in the reporting process. We have not 
had an opportunity to track down the exact source of this discrepancy. We have not used 
the Fish and Wildlife data because their collection is based upon regional fisheries and it 
is not easy to get them on a comparable basis over time by states. 

^This may be due to overfishing but this is hard to establish conclusively. Salmon seems 
to be particularly susceptible to natural fluctuations in numbers, partly because floods or 
freezing on the spawning grounds and changes in ocean temperatures affecting feeding 
supplies can bring about large variations in yield. Whatever the cause, the period in 
question has been a poor one. 



Regional Distribution of Agricultural Activity (2) / 375 






30 




||j Absolute upward shift 






20 




£j Relative downward shift 


' 


^J Absolute downward shift 


• l0 


1 


' I '"•'".., 




Me. Co 


t~~l~~]JilTth~e7 sTaies ^. 


Wts. 




Vo. 1 





1. Tex. N.J. Ui. S.C fid..:'.-: Wosh.Mxjss.Term. N.C Wk 


Mich. MdL 


50 
20 


| within these limits | . 


















' • - 


















oid shift 2 




Figu 


re 69. 


Net Shift in Fishing Employment, 1940-50. 











■::.: ^ 



State figures represent % of total net shift. Total net shift as % of 
incremental change: 36.9. Total net shift as % of 1940 fishing employment: 
8.0. Source: U. S. Bureau of the Census, U. S. Census of Population 1940. 
Vol. Ill, labor force data in State Table 11; and U. S. Census of 
Population 1950, Vol. II, labor force data in State Table 73. 



sible for this. In California there has been an expansion in the tuna catch, 
stimulated in part by favorable marine phenomena and in part by increased 
demand. 

The Gulf states have been influenced mainly by developments in shrimp and 
menhaden. Menhaden are used exclusively for the production of meal and oil. 
During the period under review here, the demand for this product was enor- 
mously expanded by the increased use of fish meal as a supplement in livestock 
feed. Most of the expansion in response to this demand has taken place in the 
Gulf states. 

The expansion in shrimp has resulted from three factors: the very great 
growth in its popularity as a food product; the fact that the frozen food process 
has reduced the risks and costs associated with serving the market; and the 
discovery of several major new beds in the Tortugas area off the coasts of 
Florida and Texas. 

The Maine fisheries have shown a large expansion because the ocean has been 
giving up record quantities of lobster and herring — both of which are restricted 
largely to the northern New England beds. It is hard to establish a reason for 
this expansion in resource supply. Many of these fisheries go through natural 
fluctuations that are not fully understood. It is thought, however, that the area 
is now beginning to reap the benefits of the size limitations on catch established 
some years ago. 



376 / Regional Distribution of Economic Activities, 1939-1954 

In short, resource exhaustion and discovery, resource replenishment, changing 
techniques for preserving and transporting, and changes in the composition of 
demand all combine to bring about the shifts observed. 

However small its absolute contributions, fishing employment has been a 
factor in the expansion of California, Texas, and Florida — three of the major 
growth states. The expansion in activity in upper New England runs counter 
to the general decline recorded for this area. 26 



Review of the Regional Influence of Agriculture 

In the last two chapters we have examined many facets of the shifts in agri- 
cultural activity. A review of the major characteristics of these changes may be 
in order. 

The regional influence of agriculture stems primarily from the fact that it is 
the major slow-growth activity among the ten 1 -digit employment sectors, 
recording a downward shift of 14.7% between 1939 and 1954. Significant long- 
term trends in supply and demand have reduced the relative importance of this 
sector of the economy over the years. Advancing technology has transformed 
many of the traditional agricultural functions into manufacturing and trade 
functions. It has also fostered more extended fabrication of agricultural products 
and has greatly increased worker productivity. At the same time, competition 
from synthetic and foreign substitutes has increased, and agricultural products 
as a whole do not have a high income elasticity of demand. These have been 
major factors in the net downward shifts observed in the Southeast, Plains, and 
northern Mountain regions (Figure 45). 

But although agricultural activity in the aggregate has not contributed to 
positive growth in the volume of economic activities, it has made important 
contributions to the growth of specific regions. In the Far West and southern 
Mountain states and in Kansas and Indiana, the net gain in agricultural employ- 
ment was a significant element in the upward shifts in total employment dis- 
played by these states. Similarly, the differential shifts in agricultural employ- 
ment, added to the proportionality effect, contributed significantly to the net 
decline in the states of the Deep South and in Montana. 27 Other states also 
experienced significant shifts in agricultural employment that ran counter to 
the general trend established by the other components of employment. In these 
cases it had the effect of dampening the total shift effect generated by the other 
sectors. 

We have seen that the differential shifts in agriculture come primarily from 
changes in field crop and livestock activities. Advancing agricultural technology 
associated with soybeans, hybrid corn, drought-resistant wheats, new feed 
sorghums, and cotton have all combined to generate a shift in field crop activity 
out of the Southeast and into the western states. The same was generally true of 

26 The expansion in fishing in this area is coupled with an expansion in poultry raising 
that also set the performance in agriculture squarely against the trend for this area. 
2T In this connection, compare Table 141 with Figures 58 and 45. 



Regional Distribution of Agricultural Activity (2) / 377 

the production of livestock products. The important exception here is that the 
Southeast gained significantly from an expansion of these products. However, 
since this shift into the Southeast was a part of the adustment away from labor 
intensive row-cropping, it was consistent with — and indeed, contributed to — the 
downward shift in farm employment in this region. 

We have also seen that each region's access to the inputs and markets that 
were significant for each of these agricultural systems was modified by many 
different types of influences. Changes both in the size and in the composition of 
total demand have changed the relative advantage of regions. As near as we 
can determine, this proportionality effect may have assisted the field crop and 
livestock areas vis-a-vis the dairy, poultry, and general farming areas. Institu- 
tional factors — both regional and national— have played a part; for example, 
the effect of price supports and the repeal of the oleomargarine tax. Changes in 
technology have modified the land resource requirements, resulting in shifts 
based upon the new requirements. Resource exhaustion has been an important 
factor, as have rivalry costs. We have seen that changes in one production sys- 
tem, by altering the competitive conditions of land use, have altered the location 
and character of otherwise independent production systems. 

We shall now briefly summarize the influence of agriculture in the growth 
behavior of specific regions, considering first the major growth areas. Agricul- 
ture contributed significantly to the growth performance of California and 
Arizona. It accounted for 12% of the differential shift in total employment in 
California and for 22% in Arizona. California claimed 24% of the nation's 
net upward shift in agricultural employment and 36% of the upward shift in 
the value of farm products sold. In the various sectors of agricultural activity, 
its share in the upward shifts was as follows: field crops — 22%, fruits and nuts — 
12%, vegetables — 57%, dairy products — 25%, poultry products — 15%, live- 
stock products — 12%, logging — 6%, and fishing — 20%. 

These changes have been associated with i